{ "id": 290005, "title": "🔥🔥Research Trends in Object Detection ", "url": "/competitions/tensorflow-great-barrier-reef/discussion/290005", "author_name": "Tensor Girl", "post_date": "2021-11-22T19:40:10.718000", "votes": 81, "comment_count": 20, "views": 0, "content": "

Hello everyone ,
\nI have compiled some of the recent research papers in Object detection for quicker reference.
\nMobileDets: Searching for Object Detection Architectures for Mobile Accelerators
\nContext R-CNN: Long Term Temporal Context for Per-Camera Object Detection
\nMnasFPN: Learning Latency-aware Pyramid Architecture for Object Detection on Mobile Devices
\nSwin Transformer V2: Scaling Up Capacity and Resolution\n
\nTransMix: Attend to Mix for Vision Transformers\n
\nRobust and Accurate Object Detection via Self-Knowledge Distillation\n
\nMasked Autoencoders Are Scalable Vision Learners\n
\nSome of the papers recommended by @crained
\nXp-GAN: Unsupervised Multi-object Controllable Video Generation - Seems like if you can detect and move an object in machine learning you must be able to detect it in the first place!
\nRethinking Drone-Based Search and Rescue with Aerial Person Detection - The interesting part of this is: Finally, we propose a novel postprocessing method for robust, approximate object localization: the merging of overlapping bounding boxes (MOB) algorithm. This final processing stage used in the AIR detector significantly improves its performance and usability in the face of real-world aerial SAR missions.
\nHope you found it useful

", "messages": [ { "id": 1592005, "postDate": "2021-11-22T19:40:10.720Z", "content": "

Hello everyone ,
\nI have compiled some of the recent research papers in Object detection for quicker reference.
\nMobileDets: Searching for Object Detection Architectures for Mobile Accelerators
\nContext R-CNN: Long Term Temporal Context for Per-Camera Object Detection
\nMnasFPN: Learning Latency-aware Pyramid Architecture for Object Detection on Mobile Devices
\nSwin Transformer V2: Scaling Up Capacity and Resolution\n
\nTransMix: Attend to Mix for Vision Transformers\n
\nRobust and Accurate Object Detection via Self-Knowledge Distillation\n
\nMasked Autoencoders Are Scalable Vision Learners\n
\nSome of the papers recommended by @crained
\nXp-GAN: Unsupervised Multi-object Controllable Video Generation - Seems like if you can detect and move an object in machine learning you must be able to detect it in the first place!
\nRethinking Drone-Based Search and Rescue with Aerial Person Detection - The interesting part of this is: Finally, we propose a novel postprocessing method for robust, approximate object localization: the merging of overlapping bounding boxes (MOB) algorithm. This final processing stage used in the AIR detector significantly improves its performance and usability in the face of real-world aerial SAR missions.
\nHope you found it useful

", "rawMarkdown": "\nHello everyone ,\n\nI have compiled some of the recent research papers in Object detection for quicker reference.\n\n\n[MobileDets: Searching for Object Detection Architectures for Mobile Accelerators](https://openaccess.thecvf.com/content/CVPR2021/papers/Xiong_MobileDets_Searching_for_Object_Detection_Architectures_for_Mobile_Accelerators_CVPR_2021_paper.pdf)\n\n[Context R-CNN: Long Term Temporal Context for Per-Camera Object Detection](https://openaccess.thecvf.com/content_CVPR_2020/papers/Beery_Context_R-CNN_Long_Term_Temporal_Context_for_Per-Camera_Object_Detection_CVPR_2020_paper.pdf)\n\n[MnasFPN: Learning Latency-aware Pyramid Architecture for Object Detection on Mobile Devices](https://arxiv.org/pdf/1912.01106v2.pdf)\n\n[Swin Transformer V2: Scaling Up Capacity and Resolution\n](https://arxiv.org/pdf/2111.09883v1.pdf)\n\n[TransMix: Attend to Mix for Vision Transformers\n](https://arxiv.org/pdf/2111.09833v1.pdf)\n\n[Robust and Accurate Object Detection via Self-Knowledge Distillation\n](https://arxiv.org/pdf/2111.07239v1.pdf)\n\n[Masked Autoencoders Are Scalable Vision Learners\n](https://arxiv.org/pdf/2111.06377v1.pdf)\n\nSome of the papers recommended by @crained \n\n\n[Xp-GAN: Unsupervised Multi-object Controllable Video Generation ](https://arxiv.org/pdf/2111.10233.pdf)- Seems like if you can detect and move an object in machine learning you must be able to detect it in the first place!\n\n[Rethinking Drone-Based Search and Rescue with Aerial Person Detection](https://arxiv.org/pdf/2111.09406.pdf) - The interesting part of this is: Finally, we propose a novel postprocessing method for robust, approximate object localization: the merging of overlapping bounding boxes (MOB) algorithm. This final processing stage used in the AIR detector significantly improves its performance and usability in the face of real-world aerial SAR missions.\n\n Hope you found it useful", "votes": 81 }, { "id": 1592107, "postDate": "2021-11-22T22:31:31.430Z", "content": "

Nice! I was just going to post. Beat me to it @usharengaraju 😄
\nHere are a few more:

\n", "rawMarkdown": "Nice! I was just going to post. Beat me to it @usharengaraju 😄\nHere are a few more:\n- [Xp-GAN: Unsupervised Multi-object Controllable Video Generation](https://arxiv.org/abs/2111.10233) - Seems like if you can detect and move an object in machine learning you must be able to detect it in the first place!\n- [Rethinking Drone-Based Search and Rescue with Aerial Person Detection](https://arxiv.org/abs/2111.09406) - The interesting part of this is: Finally, we propose a novel postprocessing method for robust, approximate object localization: the merging of overlapping bounding boxes (MOB) algorithm. This final processing stage used in the AIR detector significantly improves its performance and usability in the face of real-world aerial SAR missions.", "votes": 7, "replies": [ { "id": 1592119, "postDate": "2021-11-22T22:47:19.747Z", "content": "

@crained Ha Ha.. Thanks for sharing . Updated the post with both the papers

", "rawMarkdown": "@crained Ha Ha.. Thanks for sharing . Updated the post with both the papers" } ] }, { "id": 1662203, "postDate": "2022-01-24T05:49:46.553Z", "content": "

good job!Object Detection has been developing～

", "rawMarkdown": "good job!Object Detection has been developing～", "votes": 1 }, { "id": 1661680, "postDate": "2022-01-23T16:02:47.400Z", "content": "

Wow, Amazing!!!! Thank you for sharing the useful information!!

", "rawMarkdown": "Wow, Amazing!!!! Thank you for sharing the useful information!!", "votes": 1 }, { "id": 1594312, "postDate": "2021-11-24T17:39:59.573Z", "content": "

Thanks for sharing @usharengaraju, very useful!

", "rawMarkdown": "Thanks for sharing @usharengaraju, very useful!", "votes": 1 }, { "id": 1592603, "postDate": "2021-11-23T07:46:27.473Z", "content": "

There are a lot of models to try! Thanks for sharing!

", "rawMarkdown": "There are a lot of models to try! Thanks for sharing!", "votes": 1 }, { "id": 1692639, "postDate": "2022-02-16T06:51:41.273Z", "content": "

It is not very helpful to paste some random links. I will really upvote will all my heart if you bring out some insight or summarize even a single paper out of these. Anyway kudos to your bot army!!!

", "rawMarkdown": "It is not very helpful to paste some random links. I will really upvote will all my heart if you bring out some insight or summarize even a single paper out of these. Anyway kudos to your bot army!!!", "votes": -1 }, { "id": 1612959, "postDate": "2021-12-09T13:50:53.997Z", "rawMarkdown": "", "votes": 1, "isDeleted": true }, { "id": 1671592, "postDate": "2022-02-01T15:29:13.833Z", "content": "

Thanks for sharing @usharengaraju

", "rawMarkdown": "Thanks for sharing @usharengaraju ", "votes": 1 }, { "id": 1617244, "postDate": "2021-12-14T01:31:16.240Z", "content": "

Thanks for sharing @usharengaraju 👍

", "rawMarkdown": "Thanks for sharing @usharengaraju 👍", "votes": 1 }, { "id": 1597180, "postDate": "2021-11-27T09:45:32.767Z", "content": "

Thanks for sharing @usharengaraju 👍

", "rawMarkdown": "Thanks for sharing @usharengaraju 👍\n\n", "votes": 1 }, { "id": 1597148, "postDate": "2021-11-27T09:00:35.147Z", "content": "

Thanks for sharing @usharengaraju!

", "rawMarkdown": "Thanks for sharing @usharengaraju!", "votes": 1 }, { "id": 1596871, "postDate": "2021-11-26T23:36:37.107Z", "content": "

Thanks very mush.

", "rawMarkdown": "Thanks very mush.", "votes": 1 }, { "id": 1594375, "postDate": "2021-11-24T18:40:03.987Z", "content": "

Thanks for sharing @usharengaraju 👍

", "rawMarkdown": "Thanks for sharing @usharengaraju 👍", "votes": 1 }, { "id": 1593607, "postDate": "2021-11-24T05:12:08.323Z", "content": "

Thanks for sharing 👍

", "rawMarkdown": "Thanks for sharing 👍", "votes": 1 }, { "id": 1593482, "postDate": "2021-11-24T01:35:50.963Z", "content": "

thank you. upvoted

", "rawMarkdown": "thank you. upvoted", "votes": 1 }, { "id": 1592860, "postDate": "2021-11-23T12:25:26.620Z", "content": "

Thanks for sharing!! very helpful!!

", "rawMarkdown": "Thanks for sharing!! very helpful!!", "votes": 1 }, { "id": 1592694, "postDate": "2021-11-23T09:09:41.913Z", "content": "

Very helpful. Thanks

", "rawMarkdown": "Very helpful. Thanks", "votes": 1 }, { "id": 1592312, "postDate": "2021-11-23T03:20:30.317Z", "content": "

Thanks great resource

", "rawMarkdown": "Thanks great resource", "votes": 1 }, { "id": 1594235, "postDate": "2021-11-24T16:26:39.230Z", "content": "

Thanks for sharing🙌

", "rawMarkdown": "Thanks for sharing🙌", "votes": 2 } ], "comments": [ { "id": 1592107, "author_name": "Charlie Craine", "author_url": "", "post_date": "2021-11-22T22:31:31.430000", "content": "

Nice! I was just going to post. Beat me to it @usharengaraju 😄
\nHere are a few more:

\n", "votes": 7, "replies": [ { "id": 1592119, "author_name": "Tensor Girl", "author_url": "", "post_date": "2021-11-22T22:47:19.747000", "content": "

@crained Ha Ha.. Thanks for sharing . Updated the post with both the papers

", "votes": 0, "replies": [] } ] }, { "id": 1662203, "author_name": "Tom Kerl", "author_url": "", "post_date": "2022-01-24T05:49:46.553000", "content": "

good job!Object Detection has been developing～

", "votes": 1, "replies": [] }, { "id": 1661680, "author_name": "Eugene J. Ryu", "author_url": "", "post_date": "2022-01-23T16:02:47.400000", "content": "

Wow, Amazing!!!! Thank you for sharing the useful information!!

", "votes": 1, "replies": [] }, { "id": 1594312, "author_name": "Old Monk", "author_url": "", "post_date": "2021-11-24T17:39:59.573000", "content": "

Thanks for sharing @usharengaraju, very useful!

", "votes": 1, "replies": [] }, { "id": 1592603, "author_name": "L0Z1K", "author_url": "", "post_date": "2021-11-23T07:46:27.473000", "content": "

There are a lot of models to try! Thanks for sharing!

", "votes": 1, "replies": [] }, { "id": 1692639, "author_name": "Lisa Sharapova", "author_url": "", "post_date": "2022-02-16T06:51:41.273000", "content": "

It is not very helpful to paste some random links. I will really upvote will all my heart if you bring out some insight or summarize even a single paper out of these. Anyway kudos to your bot army!!!

", "votes": -1, "replies": [] }, { "id": 1612959, "author_name": "", "author_url": "", "post_date": "2021-12-09T13:50:53.997000", "content": "", "votes": 1, "replies": [] }, { "id": 1671592, "author_name": "SanjI", "author_url": "", "post_date": "2022-02-01T15:29:13.833000", "content": "

Thanks for sharing @usharengaraju

", "votes": 1, "replies": [] }, { "id": 1617244, "author_name": "hayaroby", "author_url": "", "post_date": "2021-12-14T01:31:16.240000", "content": "

Thanks for sharing @usharengaraju 👍

", "votes": 1, "replies": [] }, { "id": 1597180, "author_name": "Abdelrady .M", "author_url": "", "post_date": "2021-11-27T09:45:32.767000", "content": "

Thanks for sharing @usharengaraju 👍

", "votes": 1, "replies": [] }, { "id": 1597148, "author_name": "TrickyJustice", "author_url": "", "post_date": "2021-11-27T09:00:35.147000", "content": "

Thanks for sharing @usharengaraju!

", "votes": 1, "replies": [] }, { "id": 1596871, "author_name": "yangala", "author_url": "", "post_date": "2021-11-26T23:36:37.107000", "content": "

Thanks very mush.

", "votes": 1, "replies": [] }, { "id": 1594375, "author_name": "Levent Serinol", "author_url": "", "post_date": "2021-11-24T18:40:03.987000", "content": "

Thanks for sharing @usharengaraju 👍

", "votes": 1, "replies": [] }, { "id": 1593607, "author_name": "Dhinahar P", "author_url": "", "post_date": "2021-11-24T05:12:08.323000", "content": "

Thanks for sharing 👍

", "votes": 1, "replies": [] }, { "id": 1593482, "author_name": "Borna B", "author_url": "", "post_date": "2021-11-24T01:35:50.963000", "content": "

thank you. upvoted

", "votes": 1, "replies": [] }, { "id": 1592860, "author_name": "mugityan", "author_url": "", "post_date": "2021-11-23T12:25:26.620000", "content": "

Thanks for sharing!! very helpful!!

", "votes": 1, "replies": [] }, { "id": 1592694, "author_name": "Adeel Zafar", "author_url": "", "post_date": "2021-11-23T09:09:41.913000", "content": "

Very helpful. Thanks

", "votes": 1, "replies": [] }, { "id": 1592312, "author_name": "Yotoro", "author_url": "", "post_date": "2021-11-23T03:20:30.317000", "content": "

Thanks great resource

", "votes": 1, "replies": [] }, { "id": 1594235, "author_name": "ASHISH GOYAL", "author_url": "", "post_date": "2021-11-24T16:26:39.230000", "content": "

Thanks for sharing🙌

", "votes": 2, "replies": [] } ], "raw_markdown_by_id": { "1592005": "\nHello everyone ,\n\nI have compiled some of the recent research papers in Object detection for quicker reference.\n\n\n[MobileDets: Searching for Object Detection Architectures for Mobile Accelerators](https://openaccess.thecvf.com/content/CVPR2021/papers/Xiong_MobileDets_Searching_for_Object_Detection_Architectures_for_Mobile_Accelerators_CVPR_2021_paper.pdf)\n\n[Context R-CNN: Long Term Temporal Context for Per-Camera Object Detection](https://openaccess.thecvf.com/content_CVPR_2020/papers/Beery_Context_R-CNN_Long_Term_Temporal_Context_for_Per-Camera_Object_Detection_CVPR_2020_paper.pdf)\n\n[MnasFPN: Learning Latency-aware Pyramid Architecture for Object Detection on Mobile Devices](https://arxiv.org/pdf/1912.01106v2.pdf)\n\n[Swin Transformer V2: Scaling Up Capacity and Resolution\n](https://arxiv.org/pdf/2111.09883v1.pdf)\n\n[TransMix: Attend to Mix for Vision Transformers\n](https://arxiv.org/pdf/2111.09833v1.pdf)\n\n[Robust and Accurate Object Detection via Self-Knowledge Distillation\n](https://arxiv.org/pdf/2111.07239v1.pdf)\n\n[Masked Autoencoders Are Scalable Vision Learners\n](https://arxiv.org/pdf/2111.06377v1.pdf)\n\nSome of the papers recommended by @crained \n\n\n[Xp-GAN: Unsupervised Multi-object Controllable Video Generation ](https://arxiv.org/pdf/2111.10233.pdf)- Seems like if you can detect and move an object in machine learning you must be able to detect it in the first place!\n\n[Rethinking Drone-Based Search and Rescue with Aerial Person Detection](https://arxiv.org/pdf/2111.09406.pdf) - The interesting part of this is: Finally, we propose a novel postprocessing method for robust, approximate object localization: the merging of overlapping bounding boxes (MOB) algorithm. This final processing stage used in the AIR detector significantly improves its performance and usability in the face of real-world aerial SAR missions.\n\n Hope you found it useful", "1592107": "Nice! I was just going to post. Beat me to it @usharengaraju 😄\nHere are a few more:\n- [Xp-GAN: Unsupervised Multi-object Controllable Video Generation](https://arxiv.org/abs/2111.10233) - Seems like if you can detect and move an object in machine learning you must be able to detect it in the first place!\n- [Rethinking Drone-Based Search and Rescue with Aerial Person Detection](https://arxiv.org/abs/2111.09406) - The interesting part of this is: Finally, we propose a novel postprocessing method for robust, approximate object localization: the merging of overlapping bounding boxes (MOB) algorithm. This final processing stage used in the AIR detector significantly improves its performance and usability in the face of real-world aerial SAR missions.", "1662203": "good job!Object Detection has been developing～", "1661680": "Wow, Amazing!!!! Thank you for sharing the useful information!!", "1594312": "Thanks for sharing @usharengaraju, very useful!", "1592603": "There are a lot of models to try! Thanks for sharing!", "1692639": "It is not very helpful to paste some random links. I will really upvote will all my heart if you bring out some insight or summarize even a single paper out of these. Anyway kudos to your bot army!!!", "1612959": "", "1671592": "Thanks for sharing @usharengaraju ", "1617244": "Thanks for sharing @usharengaraju 👍", "1597180": "Thanks for sharing @usharengaraju 👍\n\n", "1597148": "Thanks for sharing @usharengaraju!", "1596871": "Thanks very mush.", "1594375": "Thanks for sharing @usharengaraju 👍", "1593607": "Thanks for sharing 👍", "1593482": "thank you. upvoted", "1592860": "Thanks for sharing!! very helpful!!", "1592694": "Very helpful. Thanks", "1592312": "Thanks great resource", "1594235": "Thanks for sharing🙌" } }