河川廢棄物偵測 v20260510 — YOLO11n + Ocean Cleanup
Single class debris. Train 3839 / Val 479 / Test 482. 雙 5090 DDP, 100 epochs, AdamW.
vs v20260509:
- 新增 Ocean Cleanup paper_dataset（299 imgs / 2770 bbox）。
- v509 (iwhr+tudgv only) test mAP50=0.918 / mAP50-95=0.694。
- v510 (含 ocean) test mAP50=0.886 / mAP50-95=0.654 — test 集擴大且加入 ocean 較難場景。
- 同 test set 公平比較 (v510 weights vs v509 test set): mAP50=0.921 / mAP50-95=0.691 — 持平 v509。
- 結論: v510 對舊場景表現持平，並擴展到 Ocean Cleanup 河流場景。
Test set metrics (v510 test = iwhr+tudgv+ocean)
| metric | value |
|---|
| n_images_predicted | 4141.0000 |
| map50 | 0.8863 |
| map75 | 0.7383 |
| map50_95 | 0.6545 |
| mp | 0.9053 |
| mr | 0.7841 |
| fitness | 0.6545 |
Final epoch (val) metrics
| metric | value |
|---|
| train/box_loss | 0.8488 |
| train/cls_loss | 0.4842 |
| train/dfl_loss | 0.9174 |
| metrics/precision(B) | 0.9002 |
| metrics/recall(B) | 0.8264 |
| metrics/mAP50(B) | 0.9051 |
| metrics/mAP50-95(B) | 0.6689 |
| val/box_loss | 0.9763 |
| val/cls_loss | 0.5610 |
| val/dfl_loss | 0.9620 |
Best epoch (mAP50-95) val metrics
| metric | value |
|---|
| metrics/precision(B) | 0.8892 |
| metrics/recall(B) | 0.8347 |
| metrics/mAP50(B) | 0.9090 |
| metrics/mAP50-95(B) | 0.6734 |
Train / val curves
Eval curves
PR curve
F1 curve
P curve
R curve
Confusion matrix
Confusion matrix (normalized)
Label distribution
Val batch 0 predictions
Test set sample inference (iwhr / tudgv / ocean × 2, conf≥0.25)
iwhr_1650.jpg (preds=20)
tudgv_exp23_pic263.jpg (preds=4)
ocean_1702845000_20231217_203000_PTM5077.jpg (preds=7)
iwhr_0249.jpg (preds=3)
tudgv_exp18_pic226.jpg (preds=6)
ocean_tl_ishem_ptm5116_2022-04-16T1620_1.jpg (preds=2)
Train args
task: detect
mode: train
model: /home/ubuntu/yolo11n.pt
data: /mnt/ssd/cvat2/datasets/river_debris_v510/data.yaml
epochs: 100
time: null
patience: 30
batch: 64
imgsz: 640
save: true
save_period: 10
cache: false
device: 0,1
workers: 8
project: /home/ubuntu/runs_new/river_debris_v20260510
name: run
exist_ok: true
pretrained: true
optimizer: AdamW
verbose: true
seed: 42
deterministic: true
single_cls: false
rect: false
cos_lr: true
close_mosaic: 10
resume: false
amp: true
fraction: 1.0
profile: false
freeze: null
multi_scale: 0.0
compile: false
overlap_mask: true
mask_ratio: 4
dropout: 0.0
val: true
split: val
save_json: false
conf: null
iou: 0.7
max_det: 300
half: false
dnn: false
plots: true
end2end: null
source: null
vid_stride: 1
stream_buffer: false
visualize: false
augment: false
agnostic_nms: false
classes: null
retina_masks: false
embed: null
show: false
save_frames: false
save_txt: false
save_conf: false
save_crop: false
show_labels: true
show_conf: true
show_boxes: true
line_width: null
format: torchscript
keras: false
optimize: false
int8: false
dynamic: false
simplify: true
opset: null
workspace: null
nms: false
lr0: 0.001
lrf: 0.01
momentum: 0.937
weight_decay: 0.0005
warmup_epochs: 3
warmup_momentum: 0.8
warmup_bias_lr: 0.1
box: 7.5
cls: 0.5
dfl: 1.5
pose: 12.0
kobj: 1.0
rle: 1.0
angle: 1.0
nbs: 64
hsv_h: 0.015
hsv_s: 0.7
hsv_v: 0.4
degrees: 0.0
translate: 0.1
scale: 0.5
shear: 0.0
perspective: 0.0
flipud: 0.0
fliplr: 0.5
bgr: 0.0
mosaic: 1.0
mixup: 0.0
cutmix: 0.0
copy_paste: 0.0
copy_paste_mode: flip
auto_augment: randaugment
erasing: 0.4
cfg: null
tracker: botsort.yaml
save_dir: /home/ubuntu/runs_new/river_debris_v20260510/run