RemoteCLIP

RemoteCLIP copied to clipboard

tianshanqienan

如何获得图片划分后的每个patch的表征

2

作者你好，RegoinCLIP如何获取encode_image中的的[B,N_patch,768]的结果，其中B为batch_size,N_patch为划分的patch数，谢谢🙏

Zhoumoz

hi! thank you for your excellent work. How did you overcome data bias during your training process? The number of samples in your generated data far exceeds that of RET3,...

YiguoHe

你好！ 可以提供图片到文本生成的demo代码嘛？

zzingok

data leakage issue of RSICD and RSITMD

3

The RSITMD and RSICD datasets have a data leakage issue where they might share some common images and descriptions. how to deal with it properly?

YiguoHe

Fine tuning training time?

2

I'd like to know more about how you all fine tuned your model using the base OpenCLIP weights. How long did it take and what GPUs did you end up...

BradNeuberg

Mean/Standard deviation over training dataset to remove rogue embedding dimensions outside normal magnitudes?

7

I've previously been experimenting with the RSICD CLIP model [1], which was trained just over the RSICD dataset. I'm very impressed at how many data sources you've used to train...

BradNeuberg

vitalyshalumov

Very nice work, thanks for your efforts with open source. I notice that you are combining multiple datasets to train for one large dataset. Engineering wise, how did you combine...

ranck626

feifeiwei