You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
I am very glad that someone has finally realized that keeping a image resolution of 224 or 336 is not enough to build strong VLMs for complicated vision tasks such as detection/counting 馃槃
Do you have any timeline to release your self-collected datasets?
The text was updated successfully, but these errors were encountered:
Hi @HaisongDing,
Thanks for your interest and acknowledgement in our work. Exactly we have a timeline for the release of the demo, codes and dataset. We have prepared the online demo and will release it soon after internal test and approval. Then, we'll release the dataset including the processed part from public datasets and our self-collected part. Codes and pretrained model will also be released step-by-step.
We have built a huggingface demo on our machine, but mapping from the internal network to the public network is still undergoing approval. Soon the codes and optimized data will be released.
I am very glad that someone has finally realized that keeping a image resolution of 224 or 336 is not enough to build strong VLMs for complicated vision tasks such as detection/counting 馃槃
Do you have any timeline to release your self-collected datasets?
The text was updated successfully, but these errors were encountered: