idea-research / awesome-detection-transformer Goto Github PK

Dear IDEA research team,

Thanks for organizing such fantastic repo on detection transformer.

Is it possible to add our Cascade-DETR in ICCV'23? Also, we are also interested to involve in the development of DETRX (https://github.com/IDEA-Research/detrex) if possible.

Cascade-DETR: Delving into High-Quality Universal Object Detection
Mingqiao Ye, Lei Ke, Siyuan Li, Yu-Wing Tai, Chi-Keung Tang, Martin Danelljan, Fisher Yu.
ICCV 2023. [paper] [code]

Thanks a lot.

Hi, I also recommend add Swin-transformer.
This model published in 2021 ICCV and commonly use for backbone model. Swin-transformer

Thanks to your invaluable contributions, we've developed a consolidated repository dedicated exclusively to tracking the latest advancements in transformer-based object detection methods.

https://github.com/mindgarage-shan/transformer_object_detection_survey
Feel free to check and share it!

It would be great if you could add a link to our Repository and Survey Paper.

Thanks a lot!

Hello, thanks for the nice list,

Please consider adding our recent work: https://arxiv.org/abs/2404.18849

We propose a Mixed Patches (MiPa), in conjunction with a patch-wise domain agnostic module, which is responsible for learning the best way to find a common representation of both modalities (RGB/Infrared) built on top of DINO (https://github.com/IDEA-Research/DINO).

We have a link for the code, but the work is under review, so we are waiting to release it soon here: https://github.com/heitorrapela/MiPa

Hi,

Thanks for this great repo. A paper I particularly liked is YOLOS: https://github.com/hustvl/YOLOS.

It basically shows that you can replace the convolutional backbone + encoder-decoder Transformer by just a single Transformer encoder.

Maybe you can add it as well? :)

Thank you for sharing this collection of papers

I also made a paper collection list about vision attention and transformer:
https://github.com/cmhungsteve/Awesome-Transformer-Attention

Feel free to check and share it!

I will also be appreciative if you can add a link to my repo.
Thank you

Hi! Both works use kernels (simular to object query) to build unified and end-to-end segmentation models for images and video.
K-Net, NeurIPS 2021. https://arxiv.org/abs/2106.14855
Video K-Net, CVPR-2022. https://arxiv.org/abs/2204.04656

Paper is here:
Recurrent Glimpse-based Decoder for Detection with Transformer
https://arxiv.org/abs/2112.04632

Code is here:
https://github.com/zhechen/Deformable-DETR-REGO

Hi, Thanks for good repo.

I recommend to add pyramid vision transformer(PVT) which was published in 2021 ICCV PVT repo.

And there is also PVTv2 which is PVT version2.

Have a nice day :)