search

cwq159 / PyTorch-Spiking-YOLOv3

A PyTorch implementation of Spiking-YOLOv3. Two branches are provided, based on two common PyTorch implementation of YOLOv3(ultralytics/yolov3 & eriklindernoren/PyTorch-YOLOv3), with support for Spiking-YOLOv3-Tiny at present.
GNU General Public License v3.0
207 stars 58 forks source link

About cocoapi #33

Open ryikl opened 2 years ago

ryikl commented 2 years ago

Hello, I want to change the coco dataset into my own dataset. In addition to changing labels, annotations, 5K, etc. into coco format, what else do I need to do with the coco API.

Look forward to your reply

cwq159 commented 2 years ago

No need to use the coco API. Just load your own dataset to train. Don't forget to cluster the anchors for your data.

ryikl commented 2 years ago

thank you! As you said, the ANN network is successfully trained, but the accuracy is almost 0. What is the general reason for this? What do I need to do? My image channel is 1 and the amount of data is 300.

Look forward to your reply

cwq159 commented 2 years ago

You can try to enlarge the timesteps for higher accuracy.

ryikl commented 2 years ago

Thank you for your above answer. I want to train the yolov5 model, but I only found yolov5.yaml. I directly put it in cfg/, and an error occurs when running: 'no such file or directory:' cfg/ yolov5. yaml. cfg '. Where can I find yolov5.cfg? Or is there a way to convert yaml files into cfg files?

Look forward to your reply