Pipeline

Dashboard Videos Raw frames Stills Sequences Assembly Lines

Loading sequence…

Workspaces

azad0 frames · 2 seq · 0 lines
configurator0 frames · 4 seq · 0 lines
vestas0 frames · 20 seq · 1 line
tpi0 frames · 10 seq · 0 lines
vita-demo0 frames · 3 seq · 1 line
counting0 frames · 7 seq · 1 line
Demo Space10 frames · 4 seq · 3 lines
Clock-K30 frames · 26 seq · 7 lines
Bhuvan-Test0 frames · 15 seq · 4 lines
TubeID0 frames · 8 seq · 7 lines
Cyient0 frames · 14 seq · 6 lines
True Value Hub1 frames · 1 seq · 1 line
Sanad0 frames · 1 seq · 1 line
Microplastics3 frames · 12 seq · 6 lines
Warehouse6 frames · 10 seq · 4 lines

Dev User

Settings

Sequences

ae-dot-line-n
dacha1
dacha2
dacha3
fielder
ks-dot-line-n
ks-dot-line-s
laser-v1
left-right-corner
line-dot-v3

Sequence

dacha3

ID b8438f21-9a9f-443a-af20-1b835f73559d

Updated 12 days ago

Sequence type:
Capture modality:: blank
Root:: manual

Training pipeline

Dataset

YOLO zip (data.yaml, train/images, valid/images). Use the upload icon in the card header.

Training

mAP 0.995 · 12 days ago

Metrics

time: 487.705lr/pg0: 0.000lr/pg1: 0.000lr/pg2: 0.000metrics/mAP50(B): 0.995metrics/mAP50(M): 0.995metrics/recall(B): 1.000metrics/recall(M): 1.000

ONNX imgsz

Extracts best.pt and converts to ONNX via Ultralytics (requires Python + ultralytics on this machine).

Conversion

— · 12 days ago

Core ML package ready for deployment.

Re-convert with a different input size to replace the existing CoreML package.

CoreML imgsz