mmaaz60 / mvits_for_class_agnostic_od

[ECCV'22] Official repository of paper titled "Class-agnostic Object Detection with Multi-modal Transformer".
MIT License
299 stars 24 forks source link

Applications on custom datasets are expected #3

Closed Breeze-Zero closed 2 years ago

Breeze-Zero commented 2 years ago

Now I'm not sure what data format I need to prepare to train this method

mmaaz60 commented 2 years ago

Hi @834799106,

Thank you for your interest in our work. Currently, we have only released the evaluation code along with the pre-trained models. We will be releasing the training instructions/code in a month or so.

Furthermore, as shown in Table 2 of the paper, MDef-DETR works well on many off-the-shelf datasets. You can try generating class-agnostic detections on your custom dataset and it's likely that you will get reasonable results.

nikky4D commented 2 years ago

Hi @mmaaz60 , is there any update on the training code status?

mmaaz60 commented 2 years ago

Hi @mmaaz60 , is there any update on the training code status?

Hi @nikky4D, we are planning to release the code by the end of this month. Thank you for your patience and stay tuned!

nikky4D commented 2 years ago

Any update on the training code status

mmaaz60 commented 2 years ago

Any update on the training code status

Hi @nikky4D, the code is now released. Please have a look at https://github.com/mmaaz60/mvits_for_class_agnostic_od/releases/tag/v1.0