hoya012/deep_learning_object_detection

A paper list of object detection using deep learning.

deep_learning_object_detection's Issues

Add FPS in the performance table

It will be a great help for people with real time applications in mind and want a comparison between the latest papers in this area.

DAISEE

Do you have a download method for the DAISEE data set in Mainland China?

add some links

[Dynamic R-CNN] https://arxiv.org/pdf/2004.06002.pdf
[Arbitrary-Oriented Object Detection with
Circular Smooth Label] https://arxiv.org/pdf/2003.05597.pdf
[LabelEnc] https://arxiv.org/pdf/2007.03282.pdf
[Probabilistic Anchor Assignment with IoU
Prediction for Object Detection] https://arxiv.org/pdf/2007.08103.pdf
[Soft Anchor-Point Object Detection] https://arxiv.org/pdf/1911.12448.pdf
[GeoGraph] https://arxiv.org/pdf/2003.10151.pdf
[Highly Efficient Salient Object Detection with
100K Parameters] https://arxiv.org/pdf/2003.05643.pdf
[Prior-based Domain Adaptive Object Detection
for Hazy and Rainy Conditions] https://arxiv.org/pdf/1912.00070.pdf
[Cheaper Pre-training Lunch] https://arxiv.org/pdf/2004.12178.pdf
[Side-Aware Boundary Localization for More Precise Object Detection] https://arxiv.org/pdf/1912.04260.pdf
[Domain Adaptive Object Detection via
Asymmetric Tri-way Faster-RCNN] https://arxiv.org/pdf/2007.01571.pdf
[HoughNet] https://arxiv.org/pdf/2007.02355.pdf
[OS2D] https://arxiv.org/pdf/2003.06800.pdf
[Adaptive Object Detection with Dual Multi-Label Prediction] https://arxiv.org/pdf/2003.12943.pdf
[Boosting Weakly Supervised Object Detection
with Progressive Knowledge Transfer] https://arxiv.org/pdf/2007.07986.pdf
[APRICOT] https://arxiv.org/pdf/1912.08166.pdf

Codes of 'Activity Driven Weakly Supervised Object Detection' have been DELETED by the author.

Codes of 'Activity Driven Weakly Supervised Object Detection' have been DELETED by the author.
Please update.

EfficientDet-D0-D7x

Anyone has any experiment with EfficientDet-D0-D7x? I have tried with D0, D1, D4 but I faced error with the rest

Question about "paper list from 2014 to now(2019)"

Thank you for your fantastic & easy-to-read repo!

Could you tell me why some models like SSD and YOLO are written in red characters?

performance table update

hi ， I find your performance table is not update with your paper list. Mybe you forgot .

add stars to identify the importance

hi， thanks for your sharing。would you mind add some stars to your paper list to identify the importance degree。

This is issue comments

i love niuniu!

Typo in CornerNet

First of all thanks for your great works!
I found a typo in your figure.
CornetNet -> CornerNet (last row)

'r' and 't' are very close each other in keyboard :D
I've done same typo everyday, when i write "cornernet" :(

have a nice day!

Dataset png to table

If you like I can convert the dataset png file into a HTML table in README.md.

There are some wrong URL

I have found paper url from your md file and crawled every paper.

When crawling, i noticed that some paper url that attached inside your md file are wrong...

Here is paper and wrong url.

Oh, and Thanks a lot for sharing object detection list. :)

---

3번째 Object Detection Paper : [MultiBox] Scalable Object Detection using Deep Neural Networks
--> ERROR!!

5번째 Object Detection Paper : [MRCNN] Object detection via a multiregion & semantic segmentationaware CNN model
--> ERROR!!

10번째 Object Detection Paper : [Faster RCNN, RPN] Faster RCNN: Towards RealTime Object Detection with Region Proposal Networks
--> ERROR!!

36번째 Object Detection Paper : [Mask RCNN] Mask RCNN
--> ERROR!!

44번째 Object Detection Paper : [STDN] ScaleTransferrable Object Detection
--> ERROR!!

45번째 Object Detection Paper : [RefineDet] SingleShot Refinement Neural Network for Object Detection
--> ERROR!!

50번째 Object Detection Paper : [Cascade RCNN] Cascade RCNN: Delving into High Quality Object Detection
--> ERROR!

51번째 Object Detection Paper : Finding Tiny Faces in the Wild with Generative Adversarial Network
--> ERROR!!

57번째 Object Detection Paper : ZeroAnnotation Object Detection with Web Knowledge Transfer
--> ERROR!

59번째 Object Detection Paper : [PFPNet] Parallel Feature Pyramid Network for Object Detection
--> ERROR!!

62번째 Object Detection Paper : [Pelee] Pelee: A RealTime Object Detection System on Mobile Devices
--> ERROR!!

63번째 Object Detection Paper : [HKRM] Hybrid Knowledge Routed Modules for Largescale Object Detection
--> ERROR!!

64번째 Object Detection Paper : [MetaAnchor] MetaAnchor: Learning to Detect Objects with Customized Anchors
--> ERROR!!

65번째 Object Detection Paper : [SNIPER] SNIPER: Efficient MultiScale Training
--> ERROR!!

89번째 Object Detection Paper : Adapting Object Detectors via Selective CrossDomain Alignment
--> ERROR!

90번째 Object Detection Paper : Fully Quantized Network for Object Detection
--> ERROR!!

91번째 Object Detection Paper : Distilling Object Detectors with Finegrained Feature Imitation
--> ERROR!!

92번째 Object Detection Paper : Multitask SelfSupervised Object Detection via Recycling of Bounding Box Annotations
--> ERROR!!

93번째 Object Detection Paper : [ReasoningRCNN] ReasoningRCNN: Unifying Adaptive Global Reasoning into Largescale Object Detection
--> ERROR!!

96번째 Object Detection Paper : Spatialaware Graph Relation Network for Largescale Object Detection
--> ERROR!!

97번째 Object Detection Paper : [MaxpoolNMS] MaxpoolNMS: Getting Rid of NMS Bottlenecks in TwoStage Object Detectors
--> ERROR!!

98번째 Object Detection Paper : You reap what you sow: Generating High Precision Object Proposals for Weaklysupervised Object Detection
--> ERROR!!

99번째 Object Detection Paper : Object detection with locationaware deformable convolution and backward attention filtering
--> ERROR!!

115번째 Object Detection Paper : Transductive Learning for ZeroShot Object Detection
--> ERROR!!

118번째 Object Detection Paper : [DAFS] Dynamic Anchor Feature Selection for SingleShot Object Detection
--> ERROR!!

---

deep_learning_object_detection_history.PNG missing

@hoya012 deep_learning_object_detection_history.PNG cannot be displayed.

fps column

If there is one more column for FPS of each network on 2080Ti GPU(or other GPU), the table should be perfect.

open links in new tab

Thank you for the amazing work. As an enhancement, it might be good to modify the web links such that they will be opened in a new tab when clicked.

What does '++' mean in Performance table?

There are several results which contain '(07++12)'.
What does '++' mean? Is it different just '+'?

Add demo gif to README

Disclaimer: This is a bot

It looks like your repo is trending. The github_trending_videos Instgram account automatically shows the demo gifs of trending repos in Github.

Your README doesn't seem to have any demo gifs. Add one and the next time the parser runs it will pick it up and post it on its Instagram feed. If you don't want to just close this issue we won't bother you again.

Good Job!,I would like to ask about small object detection

which models are better in small object detection，Can you give me some suggestions?

Peleenet's results in voc(07+12) is wrong.

In that paper, the result of pelee training on voc0712 is 70.9, not 76.4 that first train on coco trainval35k then fine_tune on 07+12

Good Job! It is better if the links of souce code for the papers are given .

Good job! Are there similar paper list of semantic segmentation and instance segmentation?

Great review of object detection! Are there similar paper list of semantic segmentation and instance segmentation?

MaxpoolNMS

I can't find this paper:

[MaxpoolNMS] MaxpoolNMS: Getting Rid of NMS Bottlenecks in Two-Stage Object Detectors | [CVPR' 19]

Do you have a link to share please ?
Thanks.

中文版总结：https://github.com/WIKI2020/Algorithm-is-king

https://github.com/WIKI2020/Algorithm-is-king

softer-NMS CVPR'19

softer-NMS accepted to CVPR'19
https://arxiv.org/pdf/1809.08545.pdf
code: https://github.com/yihui-he/KL-Loss

Time to update~

Dear author,

It's May!

Expecting for you update

Hardware description

It'll be nice if you could include the hardware that you performed the benchmarks on.

Link broken

The link of the paper titled "On the Importance of Data Augmentation for Object Detection" under 2020 category is broken.