Jinhong Deng
Jinhong Deng
@machuofan Thanks for your quick response. I tried using the default config of EVA L in this repo, https://github.com/CVMI-Lab/CoDet/blob/main/configs/CoDet_OVLVIS_EVA_4x.yaml, It seems the model cannot train on V100 because the memory...
Thanks for your reply. I got this. By the way, one of the baselines of the paper is BoxSup. But in the paper, it is hard to find its reference....
@lyklly Do you solve this problem?
@FionaZhangJ, I use the metric provided by the code. In my experiments, the 2012 metric usually presents better stable results, especially for the one-category scenarios. I also recommend using the...
Hi, you could follow the instructions from the cvpods to train a single-domain model (source only, i,e., training on the source) and the testing.
Yes, the foggy level is 0.02. Although the code in this repo may produce better results such as ~51.0, your results are abnormal and much higher than expectation. Based on...
@RyeChianti You may need to follow and understand the VOC data structure. The trainval.txt is a text file where each line is the image filename without suffixes.
@RyeChianti How many GPUs do you use? The default configuration is tailored for the 4 GPUs.