Target detection as an important branch of computer vision, improved significantly in recent years with the theory of neural networks and in-depth hardware GPU calculation power, the world has become a hot research artificial intelligence, landing project is also the first to begin.
Throughout 2013 to 2020, from the earliest R-CNN, OverFeat later the SSD, YOLO v3 and then last year M2Det, an endless stream of new models, performance is getting better and better! This paper will summarize the 52 full performance comparison target detection model is extremely, including a complete list of paper documents.
First the point, which listed 52 target detection model (recommended collection):
The target detection over a full summary of the technical route from a well-known project on GitHub, the author is a graduate of Electrical and Computer Engineering from Seoul National University Lee hoseong, now harvested 7.3k star.
The project addresses are:
https://github.com/hoya012/deep_learning_object_detection
The technical route runs through the timeline is 2013 to early 2020, the figure summarizes the target detection during which all models representative. Plotted red part is relatively important, it is important to master model.
Update Log
It is worth mentioning that the red stone as early as the beginning of last year has issued a document to recommend about this project, the author has also been updated as of February 2020, a major update of the following:
-
2019.2: Update 3 papers
-
2019.3: update the chart and link the code
-
2019.4: Update ICLR 2019 and CVPR 2019 paper
-
2019.5: Update CVPR 2019 paper
-
2019.6: Update CVPR 2019 thesis papers and data sets
-
2019.7: Update BMVC 2019 paper and some paper ICCV 2019
-
2019.9: Update NeurIPS 2019 paper and paper ICCV 2019
-
2019.11: AAAI 2020 update some papers and other papers
-
2020.1: Update ICLR 2020 papers and other papers
The following details!
Model performance comparison table
Since different hardware (e.g., CPU, GPU, RAM, etc.), is often not accurate enough to compare FPS. Comparative more appropriate method is to measure the performance of all models at the same hardware configuration. Performance comparison results of all the above model are as follows:
From the above table, we can clearly see the different models in VOC07, VOC12, performance on COCO data set; also lists the model papers published sources.
Here are some highlights marked red model briefly.
Model papers papers
Year 2014
R-CNN
Rich feature hierarchies for accurate object detection and semantic segmentation | [CVPR' 14]
paper:
https://arxiv.org/pdf/1311.2524.pdf
Official Code Caffe:
https://github.com/rbgirshick/rcnn
OverFeat
OverFeat: Integrated Recognition, Localization and Detection using Convolutional Networks | [ICLR' 14]
paper:
https://arxiv.org/pdf/1312.6229.pdf
Official Code Torch:
https://github.com/sermanet/OverFeat
2015
Fast R-CNN
Fast R-CNN | [ICCV' 15]
paper:
https://arxiv.org/pdf/1504.08083.pdf
Official Code caffe:
https://github.com/rbgirshick/fast-rcnn
Faster R-CNN
Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks | [NIPS' 15]
paper:
https://papers.nips.cc/paper/5638-faster-r-cnn-towards-real-time-object-detection-with-region-proposal-networks.pdf
Official Code caffe:
https://github.com/rbgirshick/py-faster-rcnn
Unofficial Code tensorflow:
https://github.com/endernewton/tf-faster-rcnn
Unofficial Code pytorch:
https://github.com/jwyang/faster-rcnn.pytorch
2016
Ohema
Training Region-based Object Detectors with Online Hard Example Mining | [CVPR' 16]
paper:
https://arxiv.org/pdf/1604.03540.pdf
Official Code caffe:
https://github.com/abhi2610/ohem
YOLO v1
You Only Look Once: Unified, Real-Time Object Detection | [CVPR' 16]
paper:
https://arxiv.org/pdf/1506.02640.pdf
Official Code c:
https://pjreddie.com/darknet/yolo/
SSD
SSD: Single Shot MultiBox Detector | [ECCV' 16]
paper:
https://arxiv.org/pdf/1512.02325.pdf
Official Code caffe:
https://github.com/weiliu89/caffe/tree/ssd
Unofficial Code tensorflow:
https://github.com/balancap/SSD-Tensorflow
Unofficial Code pytorch:
https://github.com/amdegroot/ssd.pytorch
R-FCN
R-FCN: Object Detection via Region-based Fully Convolutional Networks | [NIPS' 16]
paper:
https://arxiv.org/pdf/1605.06409.pdf
Official Code caffe:
https://github.com/daijifeng001/R-FCN
Unofficial Code caffe:
https://github.com/YuwenXiong/py-R-FCN
2017
YOLO v2
YOLO9000: Better, Faster, Stronger | [CVPR' 17]
paper:
https://arxiv.org/pdf/1612.08242.pdf
Official Code c:
https://pjreddie.com/darknet/yolo/
Unofficial Code caffe:
https://github.com/quhezheng/caffe_yolo_v2
Unofficial Code tensorflow:
https://github.com/nilboy/tensorflow-yolo
Unofficial Code tensorflow:
https://github.com/sualab/object-detection-yolov2
Unofficial Code pytorch:
https://github.com/longcw/yolo2-pytorch
FPN
Feature Pyramid Networks for Object Detection | [CVPR' 17]
paper:
http://openaccess.thecvf.com/content_cvpr_2017/papers/Lin_Feature_Pyramid_Networks_CVPR_2017_paper.pdf
Unofficial Code caffe:
https://github.com/unsky/FPN
RetinaNet
Focal Loss for Dense Object Detection | [ICCV' 17]
paper:
https://arxiv.org/pdf/1708.02002.pdf
Official Code keras:
https://github.com/fizyr/keras-retinanet
Unofficial Code pytorch:
https://github.com/kuangliu/pytorch-retinanet
Unofficial Code mxnet:
https://github.com/unsky/RetinaNet
Unofficial Code tensorflow:
https://github.com/tensorflow/tpu/tree/master/models/official/retinanet
Mask R-CNN
Mask R-CNN | [ICCV' 17]
paper:
http://openaccess.thecvf.com/content_ICCV_2017/papers/He_Mask_R-CNN_ICCV_2017_paper.pdf
Official Code caffe2:
https://github.com/facebookresearch/Detectron
Unofficial Code tensorflow:
https://github.com/matterport/Mask_RCNN
Unofficial Code tensorflow:
https://github.com/CharlesShang/FastMaskRCNN
Unofficial Code pytorch:
https://github.com/multimodallearning/pytorch-mask-rcnn
2018
YOLO v3
YOLOv3: An Incremental Improvement | [arXiv' 18]
paper:
https://pjreddie.com/media/files/papers/YOLOv3.pdf
Official Code c:
https://pjreddie.com/darknet/yolo/
Unofficial Code pytorch:
https://github.com/ayooshkathuria/pytorch-yolo-v3
Unofficial Code pytorch:
https://github.com/eriklindernoren/PyTorch-YOLOv3
Unofficial Code keras:
https://github.com/qqwweee/keras-yolo3
Unofficial Code tensorflow:
https://github.com/mystic123/tensorflow-yolo-v3
RefineDet
Single-Shot Refinement Neural Network for Object Detection | [CVPR' 18]
paper:
http://openaccess.thecvf.com/content_cvpr_2018/papers/Zhang_Single-Shot_Refinement_Neural_CVPR_2018_paper.pdf
Official Code caffe:
https://github.com/sfzhang15/RefineDet
Unofficial Code chainer:
https://github.com/fukatani/RefineDet_chainer
Unofficial Code pytorch:
https://github.com/lzx1413/PytorchSSD
2019
M2Det
M2Det: A Single-Shot Object Detector based on Multi-Level Feature Pyramid Network | [AAAI' 19]
paper:
https://arxiv.org/pdf/1811.04533.pdf
Official Code pytorch:
https://github.com/qijiezhao/M2Det
2020
Spiking-YOLO
Spiking-YOLO: Spiking Neural Network for Real-time Object Detection | [AAAI' 20]
paper:
https://arxiv.org/pdf/1903.06530.pdf
Dataset thesis papers
On the same time it is also listed above disclosed model data sets commonly used: VOC, ILSVRC, COCO, as shown in the following table:
Related Articles data set for detecting a target as follows:
Online educational experience for many years worked in an education P7 is mainly engaged in data mining and AI-depth study of the idea plus qq: 2586251002