Open medicineqian opened 2 months ago
Thank you for your interest in our research.
The information you shared does not provide enough detail to identify the cause. Additionally, I recommend using the LLVIP dataset for experiments instead of the CVC-14 data. (https://github.com/sejong-rcv/INSANet)
Hi, I re-trained the model on the CVC-14 dataset, but the loss keeps increasing. I have visualized the images and labels and they are correct.![image](https://github.com/sejong-rcv/MLPD-Multi-Label-Pedestrian-Detection/assets/74084955/a02c93ba-4bf8-4a7f-b97a-d02a0ab50bb8)