search

AILab-CVC / YOLO-World

[CVPR 2024] Real-Time Open-Vocabulary Object Detection
https://www.yoloworld.cc
GNU General Public License v3.0
4.39k stars 427 forks source link

Custom dataset detection #282

Open Cho-Hong-Seok opened 4 months ago

Cho-Hong-Seok commented 4 months ago

I want to perform PCB fault detection with a yolo-world model. I set the classes with model.set_classes, but it doesn't detect objects. How can I train it to learn about PCB faults? The dataset is in roboflow.

wondervictor commented 4 months ago

Hi @Cho-Hong-Seok, could you share some samples and text prompts? Have you fine-tuned for detecting faults?

Cho-Hong-Seok commented 4 months ago

Hi @Cho-Hong-Seok, could you share some samples and text prompts? Have you fine-tuned for detecting faults?

I download my dataset in roboflow. Here is link.(https://universe.roboflow.com/pcbfaultdetection-fygfg/fault-detection-ll4am)

And I wrote PCB fault class. classes = ["missing_hole", "mouse_bite", "short', "spur", "spurious_copper", "open_circuit" ] model.set_classes(classes)

I'm waiting your help!Thx!