Skip to content

coshz/yarncv

Folders and files

NameName
Last commit message
Last commit date

Latest commit

 

History

15 Commits
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 

Repository files navigation

About YarnCV

  1. acquisition

    • camera sampling at fps=30
  2. Yarn Dataset

    • directory: "data/refined" and "data/refined-2/"; Note: label-3 is sikpped since it depends heavily on real-time;
    • num classes=4: 0=normal, 1~4=abnormal.
  3. model, ckpt and metrics

    • models: modified resnet18, modified efficientnet-b0, yarn_sim, yarn_sim2 (best acc=100)
    • ckpt: see "out/ckpt/"
    • metrics: accuracy
  4. another attempt: visual large model

    • challenge: yarn_sim2 has poor acc (~50) on img-test
    • prompt engineering:
      • Qwen-vl-max: ~75 acc
    • future work: supervised fine-tuning
      • method: peft + LoRA
      • useful platforms: bailian.console.aliyun.com
  5. related resources (urls)

    • data: see data/.gitkeep
    • ckpt: see src/recognition/out/ckpt/.gitkeep

About

yarn deflects recognition

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published