code

[lclszsdnr]

您好，请问我没有找到训练文件train_clip_vg.py。此外，我想请教一下您的这篇优异工作本质上也是利用语言信息影响视觉特征的提取对吗，类似于qrnet、VG-LAW等工作

[linhuixiao]

@lclszsdnr I'm sorry for not having had time to tidy up the code due to a series of work delays. At present, I have released all the models and code associated with this paper. Regarding your second question, yes, this paper is essentially a further exploration of language-guided visual grounding models such as TransVG++, QRNet, and VG-LAW etc..