SOLOv2
SOLOv2 copied to clipboard
SOLOv2: Dynamic, Faster and Stronger, achives 39.5mAP on coco test-dev (36 epochs result)
After training, is it possible to do inference on CPU? How about the FPS? Thanks!
Thanks for the great work! I want to know how to get the figure of mask feature behavior in the paper? Can you tell the detailed information for the code...
in 'mmdet/models/anchor_heads/solov2_head.py', function 'solov2_target_single' generate ground truth label, but I find that, when initial cate_label, the value is 0: cate_label = torch.zeros([num_grid, num_grid], dtype=torch.int64, device=device) and then update cate_label, if...
I'm not the author, but this descends from then original mmdet based SOLO code that is BSD licensed with an academic only clause.
Where can I download the training data set
How to test coco2017 test-dev
I have installed the following versions according to SOLO installation [instructions](https://github.com/WXinlong/SOLO/blob/master/docs/INSTALL.md) using the settings from [Dockerfile](https://github.com/WXinlong/SOLO/blob/master/docker/Dockerfile) requesting **pytorch 1.3** **cuda 10.1** in an Anaconda environment: According to [torchvision setup](https://github.com/pytorch/vision#Installation) I...
Using pip 20.2.2 from /usr/local/envs/solo/lib/python3.7/site-packages/pip (python 3.7) Non-user install because site-packages writeable Created temporary directory: /tmp/pip-ephem-wheel-cache-4n_xqxkl Created temporary directory: /tmp/pip-req-tracker-5595pj5z Initialized build tracking at /tmp/pip-req-tracker-5595pj5z Created build tracker: /tmp/pip-req-tracker-5595pj5z Entered...
When I test my trained weights file,I got the message: The model and loaded state dict do not match exactly unexpected key in source state_dict: backbone.conv1.weight, backbone.bn1.weight, backbone.bn1.bias, backbone.bn1.running_mean, backbone.bn1.running_var,...