From 2de7a6850994e5fb693f7181e226ad9a57cea6bd Mon Sep 17 00:00:00 2001 From: jungfrau Date: Fri, 23 Oct 2026 18:53:13 +0200 Subject: [PATCH] run 2026-10-23_v5-s-noflip-mask1: gate_failed --- .../gold_report.json | 117 ++ .../metadata.json | 71 + .../weights.pt | 3 + .../2026-10-23_v5-s-noflip-mask1/latency.json | 1 + .../2026-10-23_v5-s-noflip-mask1/metrics.json | 57 + runs/2026-10-23_v5-s-noflip-mask1/run.json | 52 +- runs/README.md | 8 + runs/clean_gold_comparison.json | 1702 ++++++++++------- 8 files changed, 1277 insertions(+), 734 deletions(-) create mode 100644 models/registry/yolo26s-seg-v5-s-noflip-mask1-2026-10-23/gold_report.json create mode 100644 models/registry/yolo26s-seg-v5-s-noflip-mask1-2026-10-23/metadata.json create mode 100644 models/registry/yolo26s-seg-v5-s-noflip-mask1-2026-10-23/weights.pt create mode 100644 runs/2026-10-23_v5-s-noflip-mask1/latency.json create mode 100644 runs/2026-10-23_v5-s-noflip-mask1/metrics.json diff --git a/models/registry/yolo26s-seg-v5-s-noflip-mask1-2026-10-23/gold_report.json b/models/registry/yolo26s-seg-v5-s-noflip-mask1-2026-10-23/gold_report.json new file mode 100644 index 0000000..d37ad0d --- /dev/null +++ b/models/registry/yolo26s-seg-v5-s-noflip-mask1-2026-10-23/gold_report.json @@ -0,0 +1,117 @@ +{ + "per_class": { + "loop_all": { + "tp": 308, + "fp": 19, + "fn": 9, + "precision": 0.9418960244648318, + "recall": 0.9716088328075709, + "f1": 0.9565217391304348, + "mean_mask_iou": 0.9229619481348076, + "n_matched": 308 + }, + "pin": { + "tp": 214, + "fp": 23, + "fn": 10, + "precision": 0.9029535864978903, + "recall": 0.9553571428571429, + "f1": 0.9284164859002169, + "mean_mask_iou": 0.9392434589975192, + "n_matched": 214 + }, + "crystal": { + "tp": 93, + "fp": 61, + "fn": 56, + "precision": 0.6038961038961039, + "recall": 0.6241610738255033, + "f1": 0.6138613861386139, + "mean_mask_iou": 0.7860551169622239, + "n_matched": 93 + }, + "loop_face": { + "tp": 274, + "fp": 52, + "fn": 21, + "precision": 0.8404907975460123, + "recall": 0.9288135593220339, + "f1": 0.8824476650563607, + "mean_mask_iou": 0.9017649910336784, + "n_matched": 274 + }, + "ice": { + "tp": 5, + "fp": 18, + "fn": 38, + "precision": 0.21739130434782608, + "recall": 0.11627906976744186, + "f1": 0.15151515151515152, + "mean_mask_iou": 0.8077960211591213, + "n_matched": 5 + }, + "needle": { + "tp": 0, + "fp": 0, + "fn": 1, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0, + "mean_mask_iou": 0.0, + "n_matched": 0 + } + }, + "overall": { + "tp": 894, + "fp": 173, + "fn": 135, + "precision": 0.837863167760075, + "recall": 0.8688046647230321, + "f1": 0.8530534351145038, + "mean_mask_iou": 0.9054766149636467, + "n_matched": 894 + }, + "dataset": { + "gold_images": 335, + "pred_images": 335, + "missing_predictions": 0, + "iou_threshold": 0.5 + }, + "environment": { + "host": "mx-spark-01", + "ultralytics": "8.4.160", + "torch": "2.14.0+cu130" + }, + "gate": { + "baseline": "/home/jungfrau/AareLC_train/models/registry/yolo26n-seg-v4-noflip-mask1-2026-10-02/gold_report.json", + "f1_delta": 0.0053, + "mask_iou_delta": -0.01, + "mask_score": 0.7867, + "mask_score_baseline": 0.7967, + "matched_only_iou_delta": -0.0064, + "class_f1_delta": { + "loop_all": -0.0061, + "pin": -0.0043 + }, + "pass_point_rule": false, + "rule": "paired bootstrap", + "bootstrap": { + "frames": 335, + "mounts": 300, + "draws": 2000, + "f1": { + "delta": 0.0053, + "lo": -0.0099, + "hi": 0.0204, + "win": 0.753 + }, + "mask_score": { + "delta": -0.01, + "lo": -0.0217, + "hi": 0.0014, + "win": 0.053 + } + }, + "pass": false + } +} diff --git a/models/registry/yolo26s-seg-v5-s-noflip-mask1-2026-10-23/metadata.json b/models/registry/yolo26s-seg-v5-s-noflip-mask1-2026-10-23/metadata.json new file mode 100644 index 0000000..683f1d9 --- /dev/null +++ b/models/registry/yolo26s-seg-v5-s-noflip-mask1-2026-10-23/metadata.json @@ -0,0 +1,71 @@ +{ + "model_id": "yolo26s-seg-v5-s-noflip-mask1-2026-10-23", + "family": "yolo26s", + "task": "segmentation", + "training_date": "2026-10-23", + "run": "2026-10-23_v5-s-noflip-mask1", + "dataset": "dataset/2026-10-22_v5", + "base_weights": "models/base/yolo26s-seg.pt", + "hyperparams": { + "epochs": 400, + "patience": 50, + "imgsz": 640, + "batch": 16, + "device": "0", + "overlap_mask": false, + "scale": 0.7, + "translate": 0.15, + "mosaic": 0.5, + "close_mosaic": 10, + "fliplr": 0, + "flipud": 0.3, + "shear": 5.0, + "perspective": 0.0, + "mask_ratio": 1 + }, + "export_formats": [ + "pt" + ], + "metrics": { + "tp": 894, + "fp": 173, + "fn": 135, + "precision": 0.837863167760075, + "recall": 0.8688046647230321, + "f1": 0.8530534351145038, + "mean_mask_iou": 0.9054766149636467, + "n_matched": 894 + }, + "gate": { + "baseline": "/home/jungfrau/AareLC_train/models/registry/yolo26n-seg-v4-noflip-mask1-2026-10-02/gold_report.json", + "f1_delta": 0.0053, + "mask_iou_delta": -0.01, + "mask_score": 0.7867, + "mask_score_baseline": 0.7967, + "matched_only_iou_delta": -0.0064, + "class_f1_delta": { + "loop_all": -0.0061, + "pin": -0.0043 + }, + "pass_point_rule": false, + "rule": "paired bootstrap", + "bootstrap": { + "frames": 335, + "mounts": 300, + "draws": 2000, + "f1": { + "delta": 0.0053, + "lo": -0.0099, + "hi": 0.0204, + "win": 0.753 + }, + "mask_score": { + "delta": -0.01, + "lo": -0.0217, + "hi": 0.0014, + "win": 0.053 + } + }, + "pass": false + } +} diff --git a/models/registry/yolo26s-seg-v5-s-noflip-mask1-2026-10-23/weights.pt b/models/registry/yolo26s-seg-v5-s-noflip-mask1-2026-10-23/weights.pt new file mode 100644 index 0000000..78a849c --- /dev/null +++ b/models/registry/yolo26s-seg-v5-s-noflip-mask1-2026-10-23/weights.pt @@ -0,0 +1,3 @@ +version https://git-lfs.github.com/spec/v1 +oid sha256:d311d4814e9020d72890762136c0abf60e54834f3684f8fb6747f20a2631d93d +size 23397149 diff --git a/runs/2026-10-23_v5-s-noflip-mask1/latency.json b/runs/2026-10-23_v5-s-noflip-mask1/latency.json new file mode 100644 index 0000000..4e91425 --- /dev/null +++ b/runs/2026-10-23_v5-s-noflip-mask1/latency.json @@ -0,0 +1 @@ +{"weights": "runs/2026-10-23_v5-s-noflip-mask1/ultralytics/train/weights/best.pt", "frames": 150, "imgsz": 640, "infer_scale": 0.5, "half": true, "ms_median": 4.74, "ms_p90": 4.85, "ms_max": 7.22, "fps_median": 211.0, "device": "0", "backend": "pytorch", "gpu": "NVIDIA GB10", "ultralytics": "8.4.160", "torch": "2.14.0+cu130"} diff --git a/runs/2026-10-23_v5-s-noflip-mask1/metrics.json b/runs/2026-10-23_v5-s-noflip-mask1/metrics.json new file mode 100644 index 0000000..5ee49aa --- /dev/null +++ b/runs/2026-10-23_v5-s-noflip-mask1/metrics.json @@ -0,0 +1,57 @@ +{ + "overall": { + "tp": 894, + "fp": 173, + "fn": 135, + "precision": 0.837863167760075, + "recall": 0.8688046647230321, + "f1": 0.8530534351145038, + "mean_mask_iou": 0.9054766149636467, + "n_matched": 894 + }, + "gate": { + "baseline": "/home/jungfrau/AareLC_train/models/registry/yolo26n-seg-v4-noflip-mask1-2026-10-02/gold_report.json", + "f1_delta": 0.0053, + "mask_iou_delta": -0.01, + "mask_score": 0.7867, + "mask_score_baseline": 0.7967, + "matched_only_iou_delta": -0.0064, + "class_f1_delta": { + "loop_all": -0.0061, + "pin": -0.0043 + }, + "pass_point_rule": false, + "rule": "paired bootstrap", + "bootstrap": { + "frames": 335, + "mounts": 300, + "draws": 2000, + "f1": { + "delta": 0.0053, + "lo": -0.0099, + "hi": 0.0204, + "win": 0.753 + }, + "mask_score": { + "delta": -0.01, + "lo": -0.0217, + "hi": 0.0014, + "win": 0.053 + } + }, + "pass": false + }, + "training": { + "epochs_run": 235, + "precision(B)": 0.69215, + "recall(B)": 0.81816, + "mAP50(B)": 0.78793, + "mAP50-95(B)": 0.66178, + "precision(M)": 0.68409, + "recall(M)": 0.82605, + "mAP50(M)": 0.78928, + "mAP50-95(M)": 0.64104 + }, + "version": "yolo26s-seg-v5-s-noflip-mask1-2026-10-23", + "measured_at": "2026-10-23T18:53:02" +} diff --git a/runs/2026-10-23_v5-s-noflip-mask1/run.json b/runs/2026-10-23_v5-s-noflip-mask1/run.json index a3a109f..773c2bd 100644 --- a/runs/2026-10-23_v5-s-noflip-mask1/run.json +++ b/runs/2026-10-23_v5-s-noflip-mask1/run.json @@ -2,18 +2,60 @@ "id": "2026-10-23_v5-s-noflip-mask1", "dataset": "dataset/2026-10-22_v5", "base_weights": "models/base/yolo26s-seg.pt", - "status": "exported", + "status": "gate_failed", "created": "2026-10-23T03:53:49", "created_by": "jungfrau@mx-spark-01", - "hyperparams": {}, - "version": null, - "metrics": {}, + "hyperparams": { + "epochs": 400, + "patience": 50, + "imgsz": 640, + "batch": 16, + "device": "0", + "overlap_mask": false, + "scale": 0.7, + "translate": 0.15, + "mosaic": 0.5, + "close_mosaic": 10, + "fliplr": 0, + "flipud": 0.3, + "shear": 5.0, + "perspective": 0.0, + "mask_ratio": 1 + }, + "version": "yolo26s-seg-v5-s-noflip-mask1-2026-10-23", + "metrics": { + "f1": 0.8530534351145038, + "mean_mask_iou": 0.9054766149636467, + "gate_pass": false, + "baseline": "/home/jungfrau/AareLC_train/models/registry/yolo26n-seg-v4-noflip-mask1-2026-10-02/gold_report.json", + "measured_at": "2026-10-23T18:53:02" + }, "history": [ { "status": "exported", "at": "2026-10-23T03:53:49", "by": "jungfrau@mx-spark-01", "note": "promoted recipe (fliplr 0, mask_ratio 1) with yolo26s on dataset v5: gate only until its beamline speed is known, Spark" + }, + { + "status": "training", + "at": "2026-10-23T03:53:49", + "by": "jungfrau@mx-spark-01", + "note": "train_run.py" + }, + { + "status": "measured", + "at": "2026-10-23T18:53:02", + "by": "jungfrau@mx-spark-01", + "note": "" + }, + { + "status": "gate_failed", + "at": "2026-10-23T18:53:02", + "by": "jungfrau@mx-spark-01", + "note": "{\"baseline\": \"/home/jungfrau/AareLC_train/models/registry/yolo26n-seg-v4-noflip-mask1-2026-10-02/gold_report.json\", \"f1_delta\": 0.0053, \"mask_iou_delta\": -0.01, \"mask_score\": 0.7867, \"mask_score_baseline\": 0.7967, \"matched_only_iou_delta\": -0.0064, \"class_f1_delta\": {\"loop_all\": -0.0061, \"pin\": -0.0043}, \"pass_point_rule\": false, \"rule\": \"paired bootstrap\", \"bootstrap\": {\"frames\": 335, \"mounts\": 300, \"draws\": 2000, \"f1\": {\"delta\": 0.0053, \"lo\": -0.0099, \"hi\": 0.0204, \"win\": 0.753}, \"mask_score\": {\"delta\": -0.01, \"lo\": -0.0217, \"hi\": 0.0014, \"win\": 0.053}}, \"pass\": false}" } - ] + ], + "host": "mx-spark-01", + "started": "2026-10-23T03:53:49" } diff --git a/runs/README.md b/runs/README.md index 5732779..24d9bce 100644 --- a/runs/README.md +++ b/runs/README.md @@ -21,6 +21,7 @@ Generated by `python tools/ledger.py render`; do not edit by hand. | 2026-10-01_v4-noflip-mask1 | promoted | dataset/2026-10-01_v4 | models/base/yolo26n-seg.pt | yolo26n-seg-v4-noflip-mask1-2026-10-02 | 0.852 | 0.891 | yes | 2026-10-02T13:25 | | 2026-10-01_v4-recipeA | gate_failed | dataset/2026-10-01_v4 | models/base/yolo26n-seg.pt | yolo26n-seg-v4-recipeA-2026-10-02 | 0.832 | 0.884 | no | 2026-10-02T04:12 | | 2026-10-22_v5-noflip-mask1 | gate_failed | dataset/2026-10-22_v5 | models/base/yolo26n-seg.pt | yolo26n-seg-v5-noflip-mask1-2026-10-23 | 0.853 | 0.906 | no | 2026-10-23T03:53 | +| 2026-10-23_v5-s-noflip-mask1 | gate_failed | dataset/2026-10-22_v5 | models/base/yolo26s-seg.pt | yolo26s-seg-v5-s-noflip-mask1-2026-10-23 | 0.853 | 0.905 | no | 2026-10-23T18:53 | States: exported → queued → training → measured → gate_passed → gate_failed → promoted → deployed (or aborted). @@ -151,3 +152,10 @@ States: exported → queued → training → measured → gate_passed → gate_f - 2026-10-22T11:40:13 · training · jungfrau@mx-spark-01 — train_run.py - 2026-10-23T03:53:38 · measured · jungfrau@mx-spark-01 - 2026-10-23T03:53:38 · gate_failed · jungfrau@mx-spark-01 — {"baseline": "/home/jungfrau/AareLC_train/models/registry/yolo26n-seg-v4-noflip-mask1-2026-10-02/gold_report.json", "f1_delta": 0.0049, "mask_iou_delta": -0.0125, "mask_score": 0.7842, "mask_score_baseline": 0.7967, "matched_only_iou_delta": -0.0063, "class_f1_delta": {"loop_all": -0.0133, "pin": 0.0121}, "pass_point_rule": false, "rule": "paired bootstrap", "bootstrap": {"frames": 335, "mounts": 300, "draws": 2000, "f1": {"delta": 0.0049, "lo": -0.009, "hi": 0.0203, "win": 0.761}, "mask_score": {"delta": -0.0125, "lo": -0.0235, "hi": -0.0014, "win": 0.009}}, "pass": false} + +## 2026-10-23_v5-s-noflip-mask1 + +- 2026-10-23T03:53:49 · exported · jungfrau@mx-spark-01 — promoted recipe (fliplr 0, mask_ratio 1) with yolo26s on dataset v5: gate only until its beamline speed is known, Spark +- 2026-10-23T03:53:49 · training · jungfrau@mx-spark-01 — train_run.py +- 2026-10-23T18:53:02 · measured · jungfrau@mx-spark-01 +- 2026-10-23T18:53:02 · gate_failed · jungfrau@mx-spark-01 — {"baseline": "/home/jungfrau/AareLC_train/models/registry/yolo26n-seg-v4-noflip-mask1-2026-10-02/gold_report.json", "f1_delta": 0.0053, "mask_iou_delta": -0.01, "mask_score": 0.7867, "mask_score_baseline": 0.7967, "matched_only_iou_delta": -0.0064, "class_f1_delta": {"loop_all": -0.0061, "pin": -0.0043}, "pass_point_rule": false, "rule": "paired bootstrap", "bootstrap": {"frames": 335, "mounts": 300, "draws": 2000, "f1": {"delta": 0.0053, "lo": -0.0099, "hi": 0.0204, "win": 0.753}, "mask_score": {"delta": -0.01, "lo": -0.0217, "hi": 0.0014, "win": 0.053}}, "pass": false} diff --git a/runs/clean_gold_comparison.json b/runs/clean_gold_comparison.json index 87e603d..84ab0a1 100644 --- a/runs/clean_gold_comparison.json +++ b/runs/clean_gold_comparison.json @@ -1,114 +1,114 @@ { - "clean_frames": 231, + "clean_frames": 281, "reports": { "beamline yolo26n-seg-overlap-false-2026-04-12": { "per_class": { "loop_all": { - "tp": 174, - "fp": 26, - "fn": 44, - "precision": 0.87, - "recall": 0.7981651376146789, - "f1": 0.8325358851674641, - "mean_mask_iou": 0.9126279113640148, - "n_matched": 174 + "tp": 218, + "fp": 34, + "fn": 46, + "precision": 0.8650793650793651, + "recall": 0.8257575757575758, + "f1": 0.8449612403100776, + "mean_mask_iou": 0.9183915172669866, + "n_matched": 218 }, "pin": { - "tp": 141, - "fp": 36, - "fn": 27, - "precision": 0.7966101694915254, - "recall": 0.8392857142857143, - "f1": 0.8173913043478261, - "mean_mask_iou": 0.948299405704783, - "n_matched": 141 + "tp": 164, + "fp": 38, + "fn": 30, + "precision": 0.8118811881188119, + "recall": 0.845360824742268, + "f1": 0.8282828282828284, + "mean_mask_iou": 0.9449884667212329, + "n_matched": 164 }, "crystal": { - "tp": 30, - "fp": 69, - "fn": 52, - "precision": 0.30303030303030304, - "recall": 0.36585365853658536, - "f1": 0.3314917127071823, - "mean_mask_iou": 0.7330447148100457, - "n_matched": 30 + "tp": 40, + "fp": 86, + "fn": 83, + "precision": 0.31746031746031744, + "recall": 0.3252032520325203, + "f1": 0.321285140562249, + "mean_mask_iou": 0.7281942214752503, + "n_matched": 40 }, "loop_face": { - "tp": 137, - "fp": 58, - "fn": 56, - "precision": 0.7025641025641025, - "recall": 0.7098445595854922, - "f1": 0.7061855670103092, - "mean_mask_iou": 0.9264985595242958, - "n_matched": 137 + "tp": 184, + "fp": 69, + "fn": 59, + "precision": 0.7272727272727273, + "recall": 0.757201646090535, + "f1": 0.7419354838709679, + "mean_mask_iou": 0.930131750718674, + "n_matched": 184 }, "ice": { "tp": 2, - "fp": 13, + "fp": 14, "fn": 11, - "precision": 0.13333333333333333, + "precision": 0.125, "recall": 0.15384615384615385, - "f1": 0.14285714285714288, + "f1": 0.13793103448275862, "mean_mask_iou": 0.944649552595864, "n_matched": 2 } }, "overall": { - "tp": 484, - "fp": 202, - "fn": 190, - "precision": 0.7055393586005831, - "recall": 0.7181008902077152, - "f1": 0.711764705882353, - "mean_mask_iou": 0.9159471404670146, - "n_matched": 484 + "tp": 608, + "fp": 241, + "fn": 229, + "precision": 0.7161366313309776, + "recall": 0.7264038231780168, + "f1": 0.7212336892052195, + "mean_mask_iou": 0.916692054939018, + "n_matched": 608 }, "dataset": { - "gold_images": 231 + "gold_images": 281 } }, "2026-09-28_alc-v1": { "per_class": { "loop_all": { - "tp": 201, - "fp": 17, - "fn": 17, - "precision": 0.9220183486238532, - "recall": 0.9220183486238532, - "f1": 0.9220183486238532, - "mean_mask_iou": 0.9014785351834897, - "n_matched": 201 + "tp": 244, + "fp": 26, + "fn": 20, + "precision": 0.9037037037037037, + "recall": 0.9242424242424242, + "f1": 0.9138576779026217, + "mean_mask_iou": 0.9103160618585375, + "n_matched": 244 }, "pin": { - "tp": 155, - "fp": 25, + "tp": 181, + "fp": 27, "fn": 13, - "precision": 0.8611111111111112, - "recall": 0.9226190476190477, - "f1": 0.8908045977011495, - "mean_mask_iou": 0.9215466993628478, - "n_matched": 155 + "precision": 0.8701923076923077, + "recall": 0.9329896907216495, + "f1": 0.900497512437811, + "mean_mask_iou": 0.9234929164091538, + "n_matched": 181 }, "crystal": { - "tp": 32, - "fp": 40, - "fn": 50, - "precision": 0.4444444444444444, - "recall": 0.3902439024390244, - "f1": 0.4155844155844156, - "mean_mask_iou": 0.7344346881059584, - "n_matched": 32 + "tp": 47, + "fp": 49, + "fn": 76, + "precision": 0.4895833333333333, + "recall": 0.3821138211382114, + "f1": 0.4292237442922375, + "mean_mask_iou": 0.712798095999229, + "n_matched": 47 }, "loop_face": { - "tp": 148, - "fp": 46, - "fn": 45, - "precision": 0.7628865979381443, - "recall": 0.7668393782383419, - "f1": 0.7648578811369509, - "mean_mask_iou": 0.8929595506267562, - "n_matched": 148 + "tp": 195, + "fp": 52, + "fn": 48, + "precision": 0.7894736842105263, + "recall": 0.8024691358024691, + "f1": 0.7959183673469388, + "mean_mask_iou": 0.9028219908231531, + "n_matched": 195 }, "ice": { "tp": 1, @@ -122,60 +122,60 @@ } }, "overall": { - "tp": 537, - "fp": 132, - "fn": 137, - "precision": 0.8026905829596412, - "recall": 0.7967359050445104, - "f1": 0.7997021593447505, - "mean_mask_iou": 0.8949304682286074, - "n_matched": 537 + "tp": 668, + "fp": 158, + "fn": 169, + "precision": 0.8087167070217918, + "recall": 0.7980884109916367, + "f1": 0.8033674082982563, + "mean_mask_iou": 0.8977574096399805, + "n_matched": 668 }, "dataset": { - "gold_images": 231 + "gold_images": 281 } }, "2026-09-28_alc-v1-800px": { "per_class": { "loop_all": { - "tp": 204, - "fp": 31, - "fn": 14, - "precision": 0.8680851063829788, - "recall": 0.9357798165137615, - "f1": 0.9006622516556291, - "mean_mask_iou": 0.9040879144105756, - "n_matched": 204 + "tp": 248, + "fp": 41, + "fn": 16, + "precision": 0.8581314878892734, + "recall": 0.9393939393939394, + "f1": 0.8969258589511755, + "mean_mask_iou": 0.9112071454475272, + "n_matched": 248 }, "pin": { - "tp": 151, - "fp": 18, - "fn": 17, - "precision": 0.893491124260355, - "recall": 0.8988095238095238, - "f1": 0.8961424332344214, - "mean_mask_iou": 0.923705544874994, - "n_matched": 151 + "tp": 176, + "fp": 20, + "fn": 18, + "precision": 0.8979591836734694, + "recall": 0.9072164948453608, + "f1": 0.9025641025641026, + "mean_mask_iou": 0.9260759853198219, + "n_matched": 176 }, "crystal": { - "tp": 37, - "fp": 44, - "fn": 45, - "precision": 0.4567901234567901, - "recall": 0.45121951219512196, - "f1": 0.4539877300613497, - "mean_mask_iou": 0.73901798491793, - "n_matched": 37 + "tp": 51, + "fp": 66, + "fn": 72, + "precision": 0.4358974358974359, + "recall": 0.4146341463414634, + "f1": 0.42500000000000004, + "mean_mask_iou": 0.7258538188619535, + "n_matched": 51 }, "loop_face": { - "tp": 140, - "fp": 45, - "fn": 53, - "precision": 0.7567567567567568, - "recall": 0.7253886010362695, - "f1": 0.7407407407407407, - "mean_mask_iou": 0.902434314290202, - "n_matched": 140 + "tp": 187, + "fp": 48, + "fn": 56, + "precision": 0.7957446808510639, + "recall": 0.7695473251028807, + "f1": 0.7824267782426779, + "mean_mask_iou": 0.9109665444783769, + "n_matched": 187 }, "ice": { "tp": 0, @@ -189,60 +189,60 @@ } }, "overall": { - "tp": 532, - "fp": 141, - "fn": 142, - "precision": 0.7904903417533432, - "recall": 0.7893175074183977, - "f1": 0.7899034892353378, - "mean_mask_iou": 0.8977404910873555, - "n_matched": 532 + "tp": 662, + "fp": 178, + "fn": 175, + "precision": 0.7880952380952381, + "recall": 0.7909199522102748, + "f1": 0.789505068574836, + "mean_mask_iou": 0.9008127402820111, + "n_matched": 662 }, "dataset": { - "gold_images": 231 + "gold_images": 281 } }, "2026-09-28_alc-v1-humanonly": { "per_class": { "loop_all": { - "tp": 201, - "fp": 18, - "fn": 17, - "precision": 0.9178082191780822, - "recall": 0.9220183486238532, - "f1": 0.919908466819222, - "mean_mask_iou": 0.9030855439632622, - "n_matched": 201 + "tp": 244, + "fp": 26, + "fn": 20, + "precision": 0.9037037037037037, + "recall": 0.9242424242424242, + "f1": 0.9138576779026217, + "mean_mask_iou": 0.9131214122937471, + "n_matched": 244 }, "pin": { - "tp": 156, - "fp": 14, + "tp": 182, + "fp": 16, "fn": 12, - "precision": 0.9176470588235294, - "recall": 0.9285714285714286, - "f1": 0.923076923076923, - "mean_mask_iou": 0.9123383278334163, - "n_matched": 156 + "precision": 0.9191919191919192, + "recall": 0.9381443298969072, + "f1": 0.9285714285714285, + "mean_mask_iou": 0.9152615958746741, + "n_matched": 182 }, "crystal": { - "tp": 39, - "fp": 47, - "fn": 43, - "precision": 0.45348837209302323, - "recall": 0.47560975609756095, - "f1": 0.4642857142857143, - "mean_mask_iou": 0.7180698464870869, - "n_matched": 39 + "tp": 56, + "fp": 65, + "fn": 67, + "precision": 0.4628099173553719, + "recall": 0.45528455284552843, + "f1": 0.4590163934426229, + "mean_mask_iou": 0.7145956909706587, + "n_matched": 56 }, "loop_face": { - "tp": 147, - "fp": 39, - "fn": 46, - "precision": 0.7903225806451613, - "recall": 0.7616580310880829, - "f1": 0.7757255936675462, - "mean_mask_iou": 0.8957137688897161, - "n_matched": 147 + "tp": 193, + "fp": 44, + "fn": 50, + "precision": 0.8143459915611815, + "recall": 0.7942386831275721, + "f1": 0.8041666666666667, + "mean_mask_iou": 0.9040561363343581, + "n_matched": 193 }, "ice": { "tp": 1, @@ -256,60 +256,60 @@ } }, "overall": { - "tp": 544, - "fp": 122, - "fn": 130, - "precision": 0.8168168168168168, - "recall": 0.8071216617210683, - "f1": 0.8119402985074627, - "mean_mask_iou": 0.8904580399948882, - "n_matched": 544 + "tp": 676, + "fp": 155, + "fn": 161, + "precision": 0.8134777376654633, + "recall": 0.8076463560334528, + "f1": 0.8105515587529976, + "mean_mask_iou": 0.8946286690747911, + "n_matched": 676 }, "dataset": { - "gold_images": 231 + "gold_images": 281 } }, "2026-09-28_alc-v1-mask1": { "per_class": { "loop_all": { - "tp": 206, - "fp": 35, - "fn": 12, - "precision": 0.8547717842323651, - "recall": 0.944954128440367, - "f1": 0.89760348583878, - "mean_mask_iou": 0.9018385137260095, - "n_matched": 206 + "tp": 250, + "fp": 44, + "fn": 14, + "precision": 0.8503401360544217, + "recall": 0.946969696969697, + "f1": 0.8960573476702508, + "mean_mask_iou": 0.9099596381748364, + "n_matched": 250 }, "pin": { - "tp": 151, + "tp": 175, "fp": 17, - "fn": 17, - "precision": 0.8988095238095238, - "recall": 0.8988095238095238, - "f1": 0.8988095238095238, - "mean_mask_iou": 0.9333344402199405, - "n_matched": 151 + "fn": 19, + "precision": 0.9114583333333334, + "recall": 0.9020618556701031, + "f1": 0.9067357512953368, + "mean_mask_iou": 0.934956000034779, + "n_matched": 175 }, "crystal": { - "tp": 25, - "fp": 29, - "fn": 57, - "precision": 0.46296296296296297, - "recall": 0.3048780487804878, - "f1": 0.36764705882352944, - "mean_mask_iou": 0.7417127737921492, - "n_matched": 25 + "tp": 39, + "fp": 36, + "fn": 84, + "precision": 0.52, + "recall": 0.3170731707317073, + "f1": 0.3939393939393939, + "mean_mask_iou": 0.7341688197877742, + "n_matched": 39 }, "loop_face": { - "tp": 157, - "fp": 41, - "fn": 36, - "precision": 0.7929292929292929, - "recall": 0.8134715025906736, - "f1": 0.8030690537084398, - "mean_mask_iou": 0.8969021843353486, - "n_matched": 157 + "tp": 206, + "fp": 45, + "fn": 37, + "precision": 0.8207171314741036, + "recall": 0.8477366255144033, + "f1": 0.8340080971659919, + "mean_mask_iou": 0.9047398580415201, + "n_matched": 206 }, "ice": { "tp": 1, @@ -323,60 +323,60 @@ } }, "overall": { - "tp": 540, - "fp": 125, - "fn": 134, - "precision": 0.8120300751879699, - "recall": 0.8011869436201781, - "f1": 0.8065720687079909, - "mean_mask_iou": 0.9017861665176957, - "n_matched": 540 + "tp": 671, + "fp": 145, + "fn": 166, + "precision": 0.8223039215686274, + "recall": 0.8016726403823178, + "f1": 0.8118572292800968, + "mean_mask_iou": 0.9046379100020938, + "n_matched": 671 }, "dataset": { - "gold_images": 231 + "gold_images": 281 } }, "2026-09-28_alc-v1-noflip": { "per_class": { "loop_all": { - "tp": 205, - "fp": 25, - "fn": 13, - "precision": 0.8913043478260869, - "recall": 0.9403669724770642, - "f1": 0.9151785714285714, - "mean_mask_iou": 0.8986512520588249, - "n_matched": 205 + "tp": 248, + "fp": 34, + "fn": 16, + "precision": 0.8794326241134752, + "recall": 0.9393939393939394, + "f1": 0.9084249084249085, + "mean_mask_iou": 0.908630646462612, + "n_matched": 248 }, "pin": { - "tp": 153, - "fp": 20, + "tp": 179, + "fp": 21, "fn": 15, - "precision": 0.884393063583815, - "recall": 0.9107142857142857, - "f1": 0.8973607038123167, - "mean_mask_iou": 0.9216030812983836, - "n_matched": 153 + "precision": 0.895, + "recall": 0.9226804123711341, + "f1": 0.9086294416243655, + "mean_mask_iou": 0.9226412784500149, + "n_matched": 179 }, "crystal": { - "tp": 33, - "fp": 34, - "fn": 49, - "precision": 0.4925373134328358, - "recall": 0.4024390243902439, - "f1": 0.44295302013422816, - "mean_mask_iou": 0.7256520050942031, - "n_matched": 33 + "tp": 42, + "fp": 48, + "fn": 81, + "precision": 0.4666666666666667, + "recall": 0.34146341463414637, + "f1": 0.3943661971830986, + "mean_mask_iou": 0.7269361818133384, + "n_matched": 42 }, "loop_face": { - "tp": 158, - "fp": 48, - "fn": 35, - "precision": 0.7669902912621359, - "recall": 0.8186528497409327, - "f1": 0.7919799498746868, - "mean_mask_iou": 0.8941600778532955, - "n_matched": 158 + "tp": 207, + "fp": 57, + "fn": 36, + "precision": 0.7840909090909091, + "recall": 0.8518518518518519, + "f1": 0.8165680473372782, + "mean_mask_iou": 0.9055127719252077, + "n_matched": 207 }, "ice": { "tp": 1, @@ -390,60 +390,60 @@ } }, "overall": { - "tp": 550, - "fp": 128, - "fn": 124, - "precision": 0.8112094395280236, - "recall": 0.8160237388724035, - "f1": 0.8136094674556212, - "mean_mask_iou": 0.8933968167822609, - "n_matched": 550 + "tp": 677, + "fp": 161, + "fn": 160, + "precision": 0.8078758949880668, + "recall": 0.8088410991636799, + "f1": 0.808358208955224, + "mean_mask_iou": 0.9001201111381992, + "n_matched": 677 }, "dataset": { - "gold_images": 231 + "gold_images": 281 } }, "2026-09-29_alc-v1-noflip-mask1": { "per_class": { "loop_all": { - "tp": 202, - "fp": 29, - "fn": 16, - "precision": 0.8744588744588745, - "recall": 0.926605504587156, - "f1": 0.8997772828507796, - "mean_mask_iou": 0.9092301744695549, - "n_matched": 202 + "tp": 246, + "fp": 35, + "fn": 18, + "precision": 0.8754448398576512, + "recall": 0.9318181818181818, + "f1": 0.9027522935779816, + "mean_mask_iou": 0.9187634867207419, + "n_matched": 246 }, "pin": { - "tp": 155, - "fp": 20, + "tp": 181, + "fp": 21, "fn": 13, - "precision": 0.8857142857142857, - "recall": 0.9226190476190477, - "f1": 0.9037900874635569, - "mean_mask_iou": 0.9396596189883837, - "n_matched": 155 + "precision": 0.8960396039603961, + "recall": 0.9329896907216495, + "f1": 0.914141414141414, + "mean_mask_iou": 0.941242647441362, + "n_matched": 181 }, "crystal": { - "tp": 39, - "fp": 54, - "fn": 43, - "precision": 0.41935483870967744, - "recall": 0.47560975609756095, - "f1": 0.44571428571428573, - "mean_mask_iou": 0.7464256199285714, - "n_matched": 39 + "tp": 54, + "fp": 73, + "fn": 69, + "precision": 0.4251968503937008, + "recall": 0.43902439024390244, + "f1": 0.43199999999999994, + "mean_mask_iou": 0.7413464057874255, + "n_matched": 54 }, "loop_face": { - "tp": 157, - "fp": 52, - "fn": 36, - "precision": 0.7511961722488039, - "recall": 0.8134715025906736, - "f1": 0.7810945273631841, - "mean_mask_iou": 0.9031501596925571, - "n_matched": 157 + "tp": 206, + "fp": 60, + "fn": 37, + "precision": 0.7744360902255639, + "recall": 0.8477366255144033, + "f1": 0.8094302554027505, + "mean_mask_iou": 0.913591640576318, + "n_matched": 206 }, "ice": { "tp": 1, @@ -457,60 +457,60 @@ } }, "overall": { - "tp": 554, - "fp": 160, - "fn": 120, - "precision": 0.7759103641456583, - "recall": 0.8219584569732937, - "f1": 0.7982708933717579, - "mean_mask_iou": 0.9045464818155112, - "n_matched": 554 + "tp": 688, + "fp": 194, + "fn": 149, + "precision": 0.780045351473923, + "recall": 0.8219832735961768, + "f1": 0.8004653868528213, + "mean_mask_iou": 0.9091790134916127, + "n_matched": 688 }, "dataset": { - "gold_images": 231 + "gold_images": 281 } }, "2026-09-29_v2-recipeA": { "per_class": { "loop_all": { - "tp": 207, - "fp": 27, - "fn": 11, - "precision": 0.8846153846153846, - "recall": 0.9495412844036697, - "f1": 0.915929203539823, - "mean_mask_iou": 0.8945866673120488, - "n_matched": 207 + "tp": 251, + "fp": 36, + "fn": 13, + "precision": 0.8745644599303136, + "recall": 0.9507575757575758, + "f1": 0.911070780399274, + "mean_mask_iou": 0.9041942119995942, + "n_matched": 251 }, "pin": { - "tp": 157, - "fp": 24, + "tp": 183, + "fp": 26, "fn": 11, - "precision": 0.8674033149171271, - "recall": 0.9345238095238095, - "f1": 0.8997134670487106, - "mean_mask_iou": 0.9213477598934785, - "n_matched": 157 + "precision": 0.8755980861244019, + "recall": 0.9432989690721649, + "f1": 0.9081885856079405, + "mean_mask_iou": 0.9246470710148307, + "n_matched": 183 }, "crystal": { - "tp": 30, - "fp": 32, - "fn": 52, - "precision": 0.4838709677419355, - "recall": 0.36585365853658536, - "f1": 0.4166666666666667, - "mean_mask_iou": 0.7243196104684244, - "n_matched": 30 + "tp": 41, + "fp": 43, + "fn": 82, + "precision": 0.4880952380952381, + "recall": 0.3333333333333333, + "f1": 0.3961352657004831, + "mean_mask_iou": 0.7375250800516058, + "n_matched": 41 }, "loop_face": { - "tp": 155, - "fp": 61, - "fn": 38, - "precision": 0.7175925925925926, - "recall": 0.8031088082901554, - "f1": 0.7579462102689486, - "mean_mask_iou": 0.8954547981168787, - "n_matched": 155 + "tp": 203, + "fp": 67, + "fn": 40, + "precision": 0.7518518518518519, + "recall": 0.8353909465020576, + "f1": 0.7914230019493178, + "mean_mask_iou": 0.9065075251146673, + "n_matched": 203 }, "ice": { "tp": 1, @@ -524,60 +524,60 @@ } }, "overall": { - "tp": 550, - "fp": 152, - "fn": 124, - "precision": 0.7834757834757835, - "recall": 0.8160237388724035, - "f1": 0.7994186046511627, - "mean_mask_iou": 0.8931553287343031, - "n_matched": 550 + "tp": 679, + "fp": 180, + "fn": 158, + "precision": 0.7904540162980209, + "recall": 0.8112305854241338, + "f1": 0.8007075471698113, + "mean_mask_iou": 0.9002975367199315, + "n_matched": 679 }, "dataset": { - "gold_images": 231 + "gold_images": 281 } }, "2026-09-29_v3-noflip-mask1": { "per_class": { "loop_all": { - "tp": 211, - "fp": 21, - "fn": 7, - "precision": 0.9094827586206896, - "recall": 0.9678899082568807, - "f1": 0.9377777777777777, - "mean_mask_iou": 0.9276084689073371, - "n_matched": 211 + "tp": 255, + "fp": 30, + "fn": 9, + "precision": 0.8947368421052632, + "recall": 0.9659090909090909, + "f1": 0.9289617486338798, + "mean_mask_iou": 0.9332806161441184, + "n_matched": 255 }, "pin": { - "tp": 159, - "fp": 12, + "tp": 185, + "fp": 13, "fn": 9, - "precision": 0.9298245614035088, - "recall": 0.9464285714285714, - "f1": 0.9380530973451328, - "mean_mask_iou": 0.9444142180273439, - "n_matched": 159 + "precision": 0.9343434343434344, + "recall": 0.9536082474226805, + "f1": 0.9438775510204083, + "mean_mask_iou": 0.9454246238879194, + "n_matched": 185 }, "crystal": { - "tp": 42, - "fp": 31, - "fn": 40, - "precision": 0.5753424657534246, - "recall": 0.5121951219512195, - "f1": 0.5419354838709677, - "mean_mask_iou": 0.7475417406972381, - "n_matched": 42 + "tp": 57, + "fp": 45, + "fn": 66, + "precision": 0.5588235294117647, + "recall": 0.4634146341463415, + "f1": 0.5066666666666667, + "mean_mask_iou": 0.7467687656079699, + "n_matched": 57 }, "loop_face": { - "tp": 172, + "tp": 220, "fp": 49, - "fn": 21, - "precision": 0.7782805429864253, - "recall": 0.8911917098445595, - "f1": 0.8309178743961353, - "mean_mask_iou": 0.9105632296311911, - "n_matched": 172 + "fn": 23, + "precision": 0.8178438661710037, + "recall": 0.9053497942386831, + "f1": 0.859375, + "mean_mask_iou": 0.9195894751912227, + "n_matched": 220 }, "ice": { "tp": 1, @@ -591,60 +591,60 @@ } }, "overall": { - "tp": 585, - "fp": 120, - "fn": 89, - "precision": 0.8297872340425532, - "recall": 0.8679525222551929, - "f1": 0.8484408992023205, - "mean_mask_iou": 0.9141474464117182, - "n_matched": 585 + "tp": 718, + "fp": 144, + "fn": 119, + "precision": 0.8329466357308585, + "recall": 0.8578255675029869, + "f1": 0.8452030606238965, + "mean_mask_iou": 0.9173272933940786, + "n_matched": 718 }, "dataset": { - "gold_images": 231 + "gold_images": 281 } }, "2026-09-29_v3-recipeA": { "per_class": { "loop_all": { - "tp": 209, - "fp": 12, - "fn": 9, - "precision": 0.9457013574660633, - "recall": 0.9587155963302753, - "f1": 0.9521640091116172, - "mean_mask_iou": 0.9057824331232236, - "n_matched": 209 + "tp": 253, + "fp": 21, + "fn": 11, + "precision": 0.9233576642335767, + "recall": 0.9583333333333334, + "f1": 0.9405204460966543, + "mean_mask_iou": 0.9137478771325965, + "n_matched": 253 }, "pin": { - "tp": 158, + "tp": 184, "fp": 26, "fn": 10, - "precision": 0.8586956521739131, - "recall": 0.9404761904761905, - "f1": 0.8977272727272727, - "mean_mask_iou": 0.9221989660515587, - "n_matched": 158 + "precision": 0.8761904761904762, + "recall": 0.9484536082474226, + "f1": 0.9108910891089108, + "mean_mask_iou": 0.9250730136130679, + "n_matched": 184 }, "crystal": { - "tp": 35, - "fp": 32, - "fn": 47, - "precision": 0.5223880597014925, - "recall": 0.4268292682926829, - "f1": 0.4697986577181208, - "mean_mask_iou": 0.7389802909285842, - "n_matched": 35 + "tp": 51, + "fp": 46, + "fn": 72, + "precision": 0.5257731958762887, + "recall": 0.4146341463414634, + "f1": 0.46363636363636357, + "mean_mask_iou": 0.7446013406334167, + "n_matched": 51 }, "loop_face": { - "tp": 167, - "fp": 46, - "fn": 26, - "precision": 0.784037558685446, - "recall": 0.8652849740932642, - "f1": 0.8226600985221675, - "mean_mask_iou": 0.8945772860275675, - "n_matched": 167 + "tp": 215, + "fp": 50, + "fn": 28, + "precision": 0.8113207547169812, + "recall": 0.8847736625514403, + "f1": 0.8464566929133858, + "mean_mask_iou": 0.9074207092375226, + "n_matched": 215 }, "ice": { "tp": 2, @@ -658,60 +658,60 @@ } }, "overall": { - "tp": 571, - "fp": 120, - "fn": 103, - "precision": 0.8263386396526773, - "recall": 0.8471810089020771, - "f1": 0.8366300366300367, - "mean_mask_iou": 0.8965864872046884, - "n_matched": 571 + "tp": 705, + "fp": 147, + "fn": 132, + "precision": 0.8274647887323944, + "recall": 0.8422939068100358, + "f1": 0.8348134991119005, + "mean_mask_iou": 0.9023233622178666, + "n_matched": 705 }, "dataset": { - "gold_images": 231 + "gold_images": 281 } }, "2026-09-30_v3-s-640px": { "per_class": { "loop_all": { - "tp": 211, - "fp": 14, - "fn": 7, - "precision": 0.9377777777777778, - "recall": 0.9678899082568807, - "f1": 0.9525959367945823, - "mean_mask_iou": 0.9169784210661013, - "n_matched": 211 + "tp": 255, + "fp": 23, + "fn": 9, + "precision": 0.9172661870503597, + "recall": 0.9659090909090909, + "f1": 0.940959409594096, + "mean_mask_iou": 0.9243454041600878, + "n_matched": 255 }, "pin": { - "tp": 159, - "fp": 10, + "tp": 185, + "fp": 13, "fn": 9, - "precision": 0.9408284023668639, - "recall": 0.9464285714285714, - "f1": 0.9436201780415431, - "mean_mask_iou": 0.9303943558595977, - "n_matched": 159 + "precision": 0.9343434343434344, + "recall": 0.9536082474226805, + "f1": 0.9438775510204083, + "mean_mask_iou": 0.9339651180078029, + "n_matched": 185 }, "crystal": { - "tp": 38, - "fp": 36, - "fn": 44, - "precision": 0.5135135135135135, - "recall": 0.4634146341463415, - "f1": 0.48717948717948717, - "mean_mask_iou": 0.7710803169294302, - "n_matched": 38 + "tp": 56, + "fp": 53, + "fn": 67, + "precision": 0.5137614678899083, + "recall": 0.45528455284552843, + "f1": 0.48275862068965514, + "mean_mask_iou": 0.7690303030738762, + "n_matched": 56 }, "loop_face": { - "tp": 169, - "fp": 42, - "fn": 24, - "precision": 0.8009478672985783, - "recall": 0.8756476683937824, - "f1": 0.8366336633663367, - "mean_mask_iou": 0.8945585028760064, - "n_matched": 169 + "tp": 218, + "fp": 46, + "fn": 25, + "precision": 0.8257575757575758, + "recall": 0.897119341563786, + "f1": 0.8599605522682446, + "mean_mask_iou": 0.9076414034966676, + "n_matched": 218 }, "ice": { "tp": 2, @@ -722,63 +722,73 @@ "f1": 0.25, "mean_mask_iou": 0.8091053257565746, "n_matched": 2 + }, + "needle": { + "tp": 0, + "fp": 1, + "fn": 0, + "precision": 0.0, + "recall": 0.0, + "f1": 0.0, + "mean_mask_iou": 0.0, + "n_matched": 0 } }, "overall": { - "tp": 579, - "fp": 103, - "fn": 95, - "precision": 0.8489736070381232, - "recall": 0.8590504451038575, - "f1": 0.8539823008849557, - "mean_mask_iou": 0.9041706374913641, - "n_matched": 579 + "tp": 716, + "fp": 137, + "fn": 121, + "precision": 0.839390386869871, + "recall": 0.8554360812425329, + "f1": 0.8473372781065089, + "mean_mask_iou": 0.9092756403326656, + "n_matched": 716 }, "dataset": { - "gold_images": 231 + "gold_images": 281 } }, "2026-10-01_v4-noflip-mask1": { "per_class": { "loop_all": { - "tp": 211, - "fp": 11, - "fn": 7, - "precision": 0.9504504504504504, - "recall": 0.9678899082568807, - "f1": 0.9590909090909091, - "mean_mask_iou": 0.9206661897480345, - "n_matched": 211 + "tp": 256, + "fp": 16, + "fn": 8, + "precision": 0.9411764705882353, + "recall": 0.9696969696969697, + "f1": 0.955223880597015, + "mean_mask_iou": 0.9335722893688293, + "n_matched": 256 }, "pin": { - "tp": 160, - "fp": 18, + "tp": 186, + "fp": 20, "fn": 8, - "precision": 0.898876404494382, - "recall": 0.9523809523809523, - "f1": 0.9248554913294796, - "mean_mask_iou": 0.933556637400468, - "n_matched": 160 + "precision": 0.9029126213592233, + "recall": 0.9587628865979382, + "f1": 0.9299999999999999, + "mean_mask_iou": 0.9402202683221472, + "n_matched": 186 }, "crystal": { - "tp": 49, - "fp": 41, - "fn": 33, - "precision": 0.5444444444444444, - "recall": 0.5975609756097561, - "f1": 0.569767441860465, - "mean_mask_iou": 0.7753470612653858, - "n_matched": 49 + "tp": 80, + "fp": 71, + "fn": 43, + "precision": 0.5298013245033113, + "recall": 0.6504065040650406, + "f1": 0.583941605839416, + "mean_mask_iou": 0.8047452782311604, + "n_matched": 80 }, "loop_face": { - "tp": 172, - "fp": 45, + "tp": 222, + "fp": 47, "fn": 21, - "precision": 0.7926267281105991, - "recall": 0.8911917098445595, - "f1": 0.8390243902439024, - "mean_mask_iou": 0.8967733433768024, - "n_matched": 172 + "precision": 0.8252788104089219, + "recall": 0.9135802469135802, + "f1": 0.8671875000000001, + "mean_mask_iou": 0.915814646963328, + "n_matched": 222 }, "ice": { "tp": 2, @@ -792,60 +802,60 @@ } }, "overall": { - "tp": 594, - "fp": 121, - "fn": 80, - "precision": 0.8307692307692308, - "recall": 0.8813056379821959, - "f1": 0.8552915766738661, - "mean_mask_iou": 0.9049501095889037, - "n_matched": 594 + "tp": 746, + "fp": 160, + "fn": 91, + "precision": 0.8233995584988962, + "recall": 0.8912783751493429, + "f1": 0.8559954102122777, + "mean_mask_iou": 0.9158708657946052, + "n_matched": 746 }, "dataset": { - "gold_images": 231 + "gold_images": 281 } }, "2026-10-01_v4-recipeA": { "per_class": { "loop_all": { - "tp": 210, - "fp": 18, - "fn": 8, - "precision": 0.9210526315789473, - "recall": 0.963302752293578, - "f1": 0.9417040358744395, - "mean_mask_iou": 0.9083802724000448, - "n_matched": 210 + "tp": 254, + "fp": 28, + "fn": 10, + "precision": 0.900709219858156, + "recall": 0.9621212121212122, + "f1": 0.9304029304029305, + "mean_mask_iou": 0.9174162777612329, + "n_matched": 254 }, "pin": { - "tp": 157, - "fp": 17, + "tp": 183, + "fp": 19, "fn": 11, - "precision": 0.9022988505747126, - "recall": 0.9345238095238095, - "f1": 0.9181286549707602, - "mean_mask_iou": 0.9336178533746317, - "n_matched": 157 + "precision": 0.905940594059406, + "recall": 0.9432989690721649, + "f1": 0.9242424242424242, + "mean_mask_iou": 0.9366127769478735, + "n_matched": 183 }, "crystal": { - "tp": 41, - "fp": 35, - "fn": 41, - "precision": 0.5394736842105263, - "recall": 0.5, - "f1": 0.5189873417721519, - "mean_mask_iou": 0.7682728543246616, - "n_matched": 41 + "tp": 70, + "fp": 50, + "fn": 53, + "precision": 0.5833333333333334, + "recall": 0.5691056910569106, + "f1": 0.5761316872427984, + "mean_mask_iou": 0.7813824544064065, + "n_matched": 70 }, "loop_face": { - "tp": 168, - "fp": 56, + "tp": 218, + "fp": 60, "fn": 25, - "precision": 0.75, - "recall": 0.8704663212435233, - "f1": 0.8057553956834532, - "mean_mask_iou": 0.8887364101367788, - "n_matched": 168 + "precision": 0.7841726618705036, + "recall": 0.897119341563786, + "f1": 0.8368522072936659, + "mean_mask_iou": 0.9021531117701876, + "n_matched": 218 }, "ice": { "tp": 2, @@ -859,17 +869,84 @@ } }, "overall": { - "tp": 578, - "fp": 133, - "fn": 96, - "precision": 0.8129395218002813, - "recall": 0.857566765578635, - "f1": 0.8346570397111913, - "mean_mask_iou": 0.8992581330375886, - "n_matched": 578 + "tp": 727, + "fp": 164, + "fn": 110, + "precision": 0.8159371492704826, + "recall": 0.8685782556750299, + "f1": 0.8414351851851851, + "mean_mask_iou": 0.904286739599412, + "n_matched": 727 }, "dataset": { - "gold_images": 231 + "gold_images": 281 + } + }, + "2026-10-22_v5-noflip-mask1": { + "per_class": { + "loop_all": { + "tp": 257, + "fp": 23, + "fn": 7, + "precision": 0.9178571428571428, + "recall": 0.9734848484848485, + "f1": 0.9448529411764706, + "mean_mask_iou": 0.9289030179153058, + "n_matched": 257 + }, + "pin": { + "tp": 184, + "fp": 12, + "fn": 10, + "precision": 0.9387755102040817, + "recall": 0.9484536082474226, + "f1": 0.9435897435897437, + "mean_mask_iou": 0.9373288617009605, + "n_matched": 184 + }, + "crystal": { + "tp": 76, + "fp": 56, + "fn": 47, + "precision": 0.5757575757575758, + "recall": 0.6178861788617886, + "f1": 0.596078431372549, + "mean_mask_iou": 0.7788673765514786, + "n_matched": 76 + }, + "loop_face": { + "tp": 219, + "fp": 46, + "fn": 24, + "precision": 0.8264150943396227, + "recall": 0.9012345679012346, + "f1": 0.8622047244094488, + "mean_mask_iou": 0.9081406106897211, + "n_matched": 219 + }, + "ice": { + "tp": 2, + "fp": 6, + "fn": 11, + "precision": 0.25, + "recall": 0.15384615384615385, + "f1": 0.1904761904761905, + "mean_mask_iou": 0.8135394456289978, + "n_matched": 2 + } + }, + "overall": { + "tp": 738, + "fp": 143, + "fn": 99, + "precision": 0.8376844494892168, + "recall": 0.8817204301075269, + "f1": 0.859138533178114, + "mean_mask_iou": 0.9090791048881155, + "n_matched": 738 + }, + "dataset": { + "gold_images": 281 } } }, @@ -1746,405 +1823,505 @@ "dataset": { "gold_images": 148 } + }, + "2026-10-22_v5-noflip-mask1": { + "per_class": { + "loop_all": { + "tp": 134, + "fp": 10, + "fn": 4, + "precision": 0.9305555555555556, + "recall": 0.9710144927536232, + "f1": 0.9503546099290779, + "mean_mask_iou": 0.8982694113423838, + "n_matched": 134 + }, + "pin": { + "tp": 106, + "fp": 7, + "fn": 9, + "precision": 0.9380530973451328, + "recall": 0.9217391304347826, + "f1": 0.9298245614035088, + "mean_mask_iou": 0.9233067898962269, + "n_matched": 106 + }, + "crystal": { + "tp": 22, + "fp": 23, + "fn": 17, + "precision": 0.4888888888888889, + "recall": 0.5641025641025641, + "f1": 0.5238095238095238, + "mean_mask_iou": 0.8016606571748839, + "n_matched": 22 + }, + "loop_face": { + "tp": 107, + "fp": 26, + "fn": 12, + "precision": 0.8045112781954887, + "recall": 0.8991596638655462, + "f1": 0.8492063492063493, + "mean_mask_iou": 0.8618765293164179, + "n_matched": 107 + }, + "ice": { + "tp": 2, + "fp": 5, + "fn": 8, + "precision": 0.2857142857142857, + "recall": 0.2, + "f1": 0.23529411764705882, + "mean_mask_iou": 0.8135394456289978, + "n_matched": 2 + } + }, + "overall": { + "tp": 371, + "fp": 71, + "fn": 50, + "precision": 0.8393665158371041, + "recall": 0.8812351543942993, + "f1": 0.8597914252607185, + "mean_mask_iou": 0.8887413014416213, + "n_matched": 371 + }, + "dataset": { + "gold_images": 148 + } } }, "active": "yolo26n-seg-v4-noflip-mask1-2026-10-02", "bootstrap_vs_active": { "beamline yolo26n-seg-overlap-false-2026-04-12": { - "frames": 231, - "mounts": 199, + "frames": 281, + "mounts": 249, "draws": 2000, "f1": { - "delta": -0.1435, - "lo": -0.1745, - "hi": -0.1099, + "delta": -0.1348, + "lo": -0.1632, + "hi": -0.1083, "win": 0.0 }, "mask_score": { - "delta": -0.1398, - "lo": -0.1715, - "hi": -0.1087, + "delta": -0.1504, + "lo": -0.1776, + "hi": -0.123, "win": 0.0 }, "per_class": { "crystal": { "f1": { - "delta": -0.2383, - "lo": -0.3379, - "hi": -0.1362, + "delta": -0.2627, + "lo": -0.3473, + "hi": -0.182, "win": 0.0 }, "mask_score": { - "delta": -0.1951, - "lo": -0.2845, - "hi": -0.1049, + "delta": -0.2866, + "lo": -0.3721, + "hi": -0.2095, "win": 0.0 } } } }, "2026-09-28_alc-v1": { - "frames": 231, - "mounts": 199, + "frames": 281, + "mounts": 249, "draws": 2000, "f1": { - "delta": -0.0556, - "lo": -0.0822, - "hi": -0.0314, + "delta": -0.0526, + "lo": -0.0753, + "hi": -0.031, "win": 0.0 }, "mask_score": { - "delta": -0.0845, - "lo": -0.1094, - "hi": -0.0621, + "delta": -0.0998, + "lo": -0.121, + "hi": -0.0791, "win": 0.0 }, "per_class": { "crystal": { "f1": { - "delta": -0.1542, - "lo": -0.2682, - "hi": -0.0369, - "win": 0.004 + "delta": -0.1547, + "lo": -0.2465, + "hi": -0.0654, + "win": 0.002 }, "mask_score": { - "delta": -0.1767, - "lo": -0.2687, - "hi": -0.0848, - "win": 0.001 + "delta": -0.251, + "lo": -0.3303, + "hi": -0.1756, + "win": 0.0 } } } }, "2026-09-28_alc-v1-800px": { - "frames": 231, - "mounts": 199, + "frames": 281, + "mounts": 249, "draws": 2000, "f1": { - "delta": -0.0654, - "lo": -0.0895, - "hi": -0.0428, + "delta": -0.0665, + "lo": -0.0879, + "hi": -0.0444, "win": 0.0 }, "mask_score": { - "delta": -0.0889, - "lo": -0.1112, - "hi": -0.0675, + "delta": -0.1038, + "lo": -0.1239, + "hi": -0.0843, "win": 0.0 }, "per_class": { "crystal": { "f1": { - "delta": -0.1158, - "lo": -0.2154, - "hi": -0.0144, - "win": 0.013 + "delta": -0.1589, + "lo": -0.2489, + "hi": -0.0783, + "win": 0.0 }, "mask_score": { - "delta": -0.1299, - "lo": -0.2095, - "hi": -0.0432, - "win": 0.001 + "delta": -0.2224, + "lo": -0.3036, + "hi": -0.1448, + "win": 0.0 } } } }, "2026-09-28_alc-v1-humanonly": { - "frames": 231, - "mounts": 199, + "frames": 281, + "mounts": 249, "draws": 2000, "f1": { - "delta": -0.0434, - "lo": -0.0671, - "hi": -0.022, + "delta": -0.0454, + "lo": -0.0668, + "hi": -0.0242, "win": 0.0 }, "mask_score": { - "delta": -0.0788, - "lo": -0.1024, - "hi": -0.0566, + "delta": -0.0938, + "lo": -0.1158, + "hi": -0.0727, "win": 0.0 }, "per_class": { "crystal": { "f1": { - "delta": -0.1055, - "lo": -0.2073, - "hi": -0.0024, - "win": 0.022 + "delta": -0.1249, + "lo": -0.2127, + "hi": -0.0367, + "win": 0.001 }, "mask_score": { - "delta": -0.1218, - "lo": -0.2044, - "hi": -0.0383, - "win": 0.002 + "delta": -0.1981, + "lo": -0.2777, + "hi": -0.121, + "win": 0.0 } } } }, "2026-09-28_alc-v1-mask1": { - "frames": 231, - "mounts": 199, + "frames": 281, + "mounts": 249, "draws": 2000, "f1": { - "delta": -0.0487, - "lo": -0.0706, - "hi": -0.0276, + "delta": -0.0441, + "lo": -0.0641, + "hi": -0.0244, "win": 0.0 }, "mask_score": { - "delta": -0.075, - "lo": -0.0954, - "hi": -0.0558, + "delta": -0.0911, + "lo": -0.1081, + "hi": -0.0731, "win": 0.0 }, "per_class": { "crystal": { "f1": { - "delta": -0.2021, - "lo": -0.3142, - "hi": -0.0883, - "win": 0.001 + "delta": -0.19, + "lo": -0.2804, + "hi": -0.1022, + "win": 0.0 }, "mask_score": { - "delta": -0.2372, - "lo": -0.3295, - "hi": -0.1481, + "delta": -0.2906, + "lo": -0.3676, + "hi": -0.2165, "win": 0.0 } } } }, "2026-09-28_alc-v1-noflip": { - "frames": 231, - "mounts": 199, + "frames": 281, + "mounts": 249, "draws": 2000, "f1": { - "delta": -0.0417, - "lo": -0.0638, - "hi": -0.0201, - "win": 0.001 + "delta": -0.0476, + "lo": -0.0686, + "hi": -0.0279, + "win": 0.0 }, "mask_score": { - "delta": -0.0685, - "lo": -0.0896, - "hi": -0.0498, + "delta": -0.0882, + "lo": -0.1081, + "hi": -0.0704, "win": 0.0 }, "per_class": { "crystal": { "f1": { - "delta": -0.1268, - "lo": -0.2231, - "hi": -0.0227, - "win": 0.006 + "delta": -0.1896, + "lo": -0.2787, + "hi": -0.1061, + "win": 0.0 }, "mask_score": { - "delta": -0.1713, - "lo": -0.2544, - "hi": -0.0859, + "delta": -0.2752, + "lo": -0.351, + "hi": -0.2017, "win": 0.0 } } } }, "2026-09-29_alc-v1-noflip-mask1": { - "frames": 231, - "mounts": 199, + "frames": 281, + "mounts": 249, "draws": 2000, "f1": { - "delta": -0.057, - "lo": -0.0791, - "hi": -0.0358, + "delta": -0.0555, + "lo": -0.0755, + "hi": -0.0366, "win": 0.0 }, "mask_score": { - "delta": -0.054, - "lo": -0.0766, - "hi": -0.0346, + "delta": -0.069, + "lo": -0.0893, + "hi": -0.0514, "win": 0.0 }, "per_class": { "crystal": { "f1": { - "delta": -0.1241, - "lo": -0.2169, - "hi": -0.034, - "win": 0.004 + "delta": -0.1519, + "lo": -0.2288, + "hi": -0.0749, + "win": 0.0 }, "mask_score": { - "delta": -0.1083, - "lo": -0.1922, - "hi": -0.0278, - "win": 0.004 + "delta": -0.1979, + "lo": -0.2752, + "hi": -0.1223, + "win": 0.0 } } } }, "2026-09-29_v2-recipeA": { - "frames": 231, - "mounts": 199, + "frames": 281, + "mounts": 249, "draws": 2000, "f1": { - "delta": -0.0559, - "lo": -0.0806, - "hi": -0.033, + "delta": -0.0553, + "lo": -0.0777, + "hi": -0.0346, "win": 0.0 }, "mask_score": { - "delta": -0.0687, - "lo": -0.0921, - "hi": -0.0489, + "delta": -0.0859, + "lo": -0.1061, + "hi": -0.0673, "win": 0.0 }, "per_class": { "crystal": { "f1": { - "delta": -0.1531, - "lo": -0.2622, - "hi": -0.0444, - "win": 0.003 + "delta": -0.1878, + "lo": -0.2776, + "hi": -0.1027, + "win": 0.0 }, "mask_score": { - "delta": -0.1983, - "lo": -0.2899, - "hi": -0.1077, + "delta": -0.2776, + "lo": -0.3567, + "hi": -0.2033, "win": 0.0 } } } }, "2026-09-29_v3-noflip-mask1": { - "frames": 231, - "mounts": 199, + "frames": 281, + "mounts": 249, "draws": 2000, "f1": { - "delta": -0.0069, - "lo": -0.0267, - "hi": 0.0122, - "win": 0.226 + "delta": -0.0108, + "lo": -0.0294, + "hi": 0.0066, + "win": 0.108 }, "mask_score": { - "delta": -0.0041, - "lo": -0.0183, - "hi": 0.0095, - "win": 0.269 - }, - "per_class": { - "crystal": { - "f1": { - "delta": -0.0278, - "lo": -0.11, - "hi": 0.0547, - "win": 0.241 - }, - "mask_score": { - "delta": -0.0804, - "lo": -0.1543, - "hi": -0.0154, - "win": 0.009 - } - } - } - }, - "2026-09-29_v3-recipeA": { - "frames": 231, - "mounts": 199, - "draws": 2000, - "f1": { - "delta": -0.0187, - "lo": -0.0387, - "hi": 0.0022, - "win": 0.039 - }, - "mask_score": { - "delta": -0.038, - "lo": -0.0534, - "hi": -0.0229, + "delta": -0.0294, + "lo": -0.0444, + "hi": -0.0143, "win": 0.0 }, "per_class": { "crystal": { "f1": { - "delta": -0.1, - "lo": -0.1933, - "hi": 0.0008, - "win": 0.025 + "delta": -0.0773, + "lo": -0.1503, + "hi": -0.0088, + "win": 0.016 }, "mask_score": { - "delta": -0.1479, - "lo": -0.228, - "hi": -0.0702, + "delta": -0.1773, + "lo": -0.2457, + "hi": -0.1106, + "win": 0.0 + } + } + } + }, + "2026-09-29_v3-recipeA": { + "frames": 281, + "mounts": 249, + "draws": 2000, + "f1": { + "delta": -0.0212, + "lo": -0.0411, + "hi": -0.0018, + "win": 0.015 + }, + "mask_score": { + "delta": -0.0563, + "lo": -0.0722, + "hi": -0.0407, + "win": 0.0 + }, + "per_class": { + "crystal": { + "f1": { + "delta": -0.1203, + "lo": -0.2096, + "hi": -0.0407, + "win": 0.001 + }, + "mask_score": { + "delta": -0.2147, + "lo": -0.2904, + "hi": -0.1485, "win": 0.0 } } } }, "2026-09-30_v3-s-640px": { - "frames": 231, - "mounts": 199, + "frames": 281, + "mounts": 249, "draws": 2000, "f1": { - "delta": -0.0013, - "lo": -0.0209, - "hi": 0.0174, - "win": 0.457 + "delta": -0.0087, + "lo": -0.027, + "hi": 0.0097, + "win": 0.172 }, "mask_score": { - "delta": -0.0208, - "lo": -0.035, - "hi": -0.0084, + "delta": -0.0385, + "lo": -0.0528, + "hi": -0.0245, "win": 0.0 }, "per_class": { "crystal": { "f1": { - "delta": -0.0826, - "lo": -0.1837, - "hi": 0.0103, - "win": 0.044 + "delta": -0.1012, + "lo": -0.1864, + "hi": -0.0154, + "win": 0.008 }, "mask_score": { - "delta": -0.106, - "lo": -0.1851, - "hi": -0.0332, - "win": 0.003 + "delta": -0.1733, + "lo": -0.2461, + "hi": -0.1046, + "win": 0.0 } } } }, "2026-10-01_v4-recipeA": { - "frames": 231, - "mounts": 199, + "frames": 281, + "mounts": 249, "draws": 2000, "f1": { - "delta": -0.0206, - "lo": -0.0373, - "hi": -0.0036, - "win": 0.006 + "delta": -0.0146, + "lo": -0.0298, + "hi": 0.001, + "win": 0.031 }, "mask_score": { - "delta": -0.0264, - "lo": -0.0398, - "hi": -0.0139, + "delta": -0.0309, + "lo": -0.0427, + "hi": -0.0189, "win": 0.0 }, "per_class": { "crystal": { "f1": { - "delta": -0.0508, - "lo": -0.1195, - "hi": 0.0216, - "win": 0.077 + "delta": -0.0078, + "lo": -0.0736, + "hi": 0.0523, + "win": 0.42 }, "mask_score": { - "delta": -0.0792, - "lo": -0.1367, - "hi": -0.0252, + "delta": -0.0787, + "lo": -0.1316, + "hi": -0.0284, "win": 0.001 } } } + }, + "2026-10-22_v5-noflip-mask1": { + "frames": 281, + "mounts": 249, + "draws": 2000, + "f1": { + "delta": 0.0031, + "lo": -0.0139, + "hi": 0.0189, + "win": 0.632 + }, + "mask_score": { + "delta": -0.0147, + "lo": -0.0275, + "hi": -0.0029, + "win": 0.009 + }, + "per_class": { + "crystal": { + "f1": { + "delta": 0.0121, + "lo": -0.0606, + "hi": 0.0802, + "win": 0.607 + }, + "mask_score": { + "delta": -0.0422, + "lo": -0.0993, + "hi": 0.0126, + "win": 0.065 + } + } + } } }, "core_reports_usable_light": { @@ -3018,6 +3195,73 @@ "dataset": { "gold_images": 142 } + }, + "2026-10-22_v5-noflip-mask1": { + "per_class": { + "loop_all": { + "tp": 132, + "fp": 10, + "fn": 4, + "precision": 0.9295774647887324, + "recall": 0.9705882352941176, + "f1": 0.9496402877697842, + "mean_mask_iou": 0.8972836658405989, + "n_matched": 132 + }, + "pin": { + "tp": 103, + "fp": 7, + "fn": 7, + "precision": 0.9363636363636364, + "recall": 0.9363636363636364, + "f1": 0.9363636363636364, + "mean_mask_iou": 0.9262646869129953, + "n_matched": 103 + }, + "crystal": { + "tp": 22, + "fp": 23, + "fn": 17, + "precision": 0.4888888888888889, + "recall": 0.5641025641025641, + "f1": 0.5238095238095238, + "mean_mask_iou": 0.8016606571748839, + "n_matched": 22 + }, + "loop_face": { + "tp": 106, + "fp": 24, + "fn": 12, + "precision": 0.8153846153846154, + "recall": 0.8983050847457628, + "f1": 0.8548387096774194, + "mean_mask_iou": 0.8625092777552028, + "n_matched": 106 + }, + "ice": { + "tp": 2, + "fp": 5, + "fn": 8, + "precision": 0.2857142857142857, + "recall": 0.2, + "f1": 0.23529411764705882, + "mean_mask_iou": 0.8135394456289978, + "n_matched": 2 + } + }, + "overall": { + "tp": 365, + "fp": 69, + "fn": 48, + "precision": 0.8410138248847926, + "recall": 0.8837772397094431, + "f1": 0.8618654073199528, + "mean_mask_iou": 0.8891405573538479, + "n_matched": 365 + }, + "dataset": { + "gold_images": 142 + } } }, "saturated_frames": [ @@ -3028,7 +3272,7 @@ "tpi_61343", "tpi_61381" ], - "mounts": 199, + "mounts": 249, "excluded": { "tpi_14688": "same image as tpi_14689 (uploaded twice), labelled differently" }