CornerNet: Detecting Objects as Paired Keypoints (Paper Explained)

Подписаться 259 тыс.

Просмотров 12 тыс.

50% 1

Many object detectors focus on locating the center of the object they want to find. However, this leaves them with the secondary problem of determining the specifications of the bounding box, leading to undesirable solutions like anchor boxes. This paper directly detects the top left and the bottom right corners of objects independently, along with descriptors that allows to match the two later and form a complete bounding box. For this, a new pooling method, called corner pooling, is introduced.
OUTLINE:
0:00 - Intro & High-Level Overview
1:40 - Object Detection
2:40 - Pipeline I - Hourglass
4:00 - Heatmap & Embedding Outputs
8:40 - Heatmap Loss
10:55 - Embedding Loss
14:35 - Corner Pooling
20:40 - Experiments
Paper: arxiv.org/abs/1808.01244
Code: github.com/princeton-vl/Corne...
Abstract:
We propose CornerNet, a new approach to object detection where we detect an object bounding box as a pair of keypoints, the top-left corner and the bottom-right corner, using a single convolution neural network. By detecting objects as paired keypoints, we eliminate the need for designing a set of anchor boxes commonly used in prior single-stage detectors. In addition to our novel formulation, we introduce corner pooling, a new type of pooling layer that helps the network better localize corners. Experiments show that CornerNet achieves a 42.2% AP on MS COCO, outperforming all existing one-stage detectors.
Authors: Hei Law, Jia Deng
Links:
RU-vid: / yannickilcher
Twitter: / ykilcher
BitChute: www.bitchute.com/channel/yann...
Minds: www.minds.com/ykilcher

Наука

Опубликовано:

2 авг 2024

Ссылка:

Скачать:

Готовим ссылку...

Добавить в:

Мой плейлист

Посмотреть позже

Комментарии : 23

@teslaonly2136 4 года назад

I almost finished 1/3 of your uploaded videos. It feels like someone is reading papers with me. The feeling is great! Thanks so much Yannic. Keep it up!

@AediWang 4 года назад

Paper from 1 year ago is now "a bit old". Just amazing how fast the field moves.

@siddharthbhargava4857 4 года назад

Thank you for the explanation.

@jasdeepsinghgrover2470 4 года назад

This shows that there still so much left in DL to be done. One thing I see is that it seems that every point is making a prediction. According to research by Uber, there are location sensitive CNNs which can also be tried in these cases. Would love to see something like a combination of the two ideas.

@herp_derpingson 4 года назад

These embeddings get more interesting the more you think about it. It is essentially two neural networks inventing a language to talk to each other. If we can make this interpretable, it might open up a lot of possibilities.

@AIwithAniket 4 года назад

Nicely explained

@ruanjiayang 2 года назад

Corner pooling is a really smart way to largely increase perception field, sort of like deformable convolution. But DETR seems to generally solve the problem in object detection since it make use of the full image as perception field.

@kwonohhyeok5016 3 года назад

first subscription in my life. thanks for your video

@rickywong8149 4 года назад

I really like your content, can you make an explanation of centernet : objects as point i dont really quite get the idea of its loss function

@awangprajaanugerah8231 2 дня назад

How can i find the research paper like you do

@austinmw89 3 года назад

Hey, have you done videos on the older but still heavily used architectures Faster RCNN, SSD, YOLO3, RetinaNet?

@larrybird3729 4 года назад

The person who put the thumbs down has Oppositional defiant disorder (ODD)🤣

@l33tc0d3 4 года назад

My intuition is that using paired keypoints is cheaper but should be more inaccurate over anchor boxes. For example, It is not clear what the paper does when there are overlapping objects that share the same keypoint (e.g. top-left). Using keypoints is interesting nevertheless. I found another recent paper that just uses keypoints inside transformer to replace RGB tracking and matching pipeline for pose tracking task: arxiv.org/pdf/1912.02323.pdf

@0lec817 4 года назад

Any specific reason you went with this approach over any of the other very similar boxless/keypoint detection approaches like CenterNet ("Objects as Points") or CSPNet ("Center and Scale Prediction") that not even require laborious embeddings while performing equally or even better? Or the "CenterNet: Keypoint Triplets for Object Detection" paper that basically is the combination of the CornerNet and the Center approaches. I mean they basically all do the same (keypoint detection) which in my opinion is quite different to what you suggested with the cross attention matrix from the attention heads?

@YannicKilcher 4 года назад

Yes this paper didn't turn out to be exactly what I hoped, but still interesting. I chose it just because it sounded like fun.

@efedoganay07 4 года назад

So, does network predict a tensor of WxHxC for heatmap branch ?

@YannicKilcher 4 года назад

Yes, one for top left and one for bottom right

@SadatAShaik 4 года назад

Your videos are great!! Keep them up :) Why do you think they decided to go with these push and pull losses instead of using a triplet loss? Seems almost identical to the push + pull losses they propose

@YannicKilcher 4 года назад

No idea, but it's either the first thing they tried, or they tried a bunch of things and this worked the best.