Wangzhen
Wangzhen
Hi, Thanks for your great work!!! I see files `captions_train2014.json` and `captions_val2014.json` under folder `SPICE/src/main/data/coco/annotations/` . In fact, I want to all reference graphs of COCO, but I'm not sure...
Hi, Thanks for sharing. But I have a question, I hope I can get the answer. We can usually get 36 detection objects using Faster R-CNN, but I see that...
你好,感谢您的复现工作,不过我使用自己的数据训练该模型,loss不会下降,请问我该如何排查原因? 我的数据为中英文均包含的干净录音,添加musan噪声后作为训练数据,使用mixloss,mixloss值稳定在40,sisnr值稳定在7~8之间,且不会下降和提升。
I got the following error, anybody has solution? Traceback (most recent call last): File "./tools/test_net.py", line 89, in imdb = get_imdb(args.roidb, args.imdb, args.rpndb, split=2, num_im=args.test_size) File "/home/jonas/桌面/scene-graph-TF-release/tools/../lib/datasets/factory.py", line 4, in...
Hi, I trained the model with my own dataset. However, the duration loss increases gradually after a period of time for val data, it's overfitting, right? Can I stop the...
Thanks your great work!!! In the paper, I see that the model will eventually generate the scene graph. How can I get the scene graph?
Hi, thank you again for your great work. Now I have another problem that I need your help. I process many graphs one by one. Because the value of `features_nonzero`...
Hello again, I want to know if the nodes in a graph are disconnected (meaning that there are paths between some of the nodes, but there are three or four...
Thank you for sharing your perfect job. But I'm not sure about the detailed process of your program generating text only from the title. How is the graph input in...
Hi, I tried to train on my dataset, but I seem to have an abnormal loss curve. Do you have any suggestions? Thanks. ----------------------------------------------------- The loss of AR: https://drive.google.com/file/d/1-gZJX-mwYZ-2vkKTl0dTwBcp1A8MHrmV/view?usp=drive_link ...