Object Detection in Videos with Tubelet Proposal...

1
Object Detection in Videos with Tubelet Proposal Networks Proposals in Video Object Detection Tubelet Proposal Networks Qualitative Results Framework Encoder-decoder LSTM YouTube-Objects Dataset Experimental Settings Kai Kang 1 , Hongsheng Li 1 , Tong Xiao 1 , Wanli Ouyang 1,4 , Junjie Yan 2 , Xihui Liu 3 , Xiaogang Wang 1 1 The Chinese University of Hong Kong, 2 SenseTime Group Limited 3 Tsinghua University, 4 The University of Sydney Results on ImageNet VID Dataset Frames Per-frame static proposals Regression to GT Boxes (RegressBox) Regression to GT Movement (RegressMove, Ours) t y x Tubelet Proposal Network Motion Prediction Encoder LSTM Decoder LSTM Class Label Tubelet Features Tubelet Proposal Network Encoder-decoder LSTM Classification CNN Spatial Anchors t y x Tubelet CNN Tubelet Generation A B A A A A B B B B 4 2f 16 5f W 2 b 2 W 5 b 5 Block Initialization Iter 1 Iter 2 Iter 3 Iter 4 Iter 5 l Parallel Generation Class Label Decoder LSTM Encoder LSTM Tubelet Features Object localization on YouTube-Objects (YTO) dataset Qualitative results on ImageNet VID validation set Results on ImageNet VID validation set Initialization of tubelet proposal networks

Transcript of Object Detection in Videos with Tubelet Proposal...

Page 1: Object Detection in Videos with Tubelet Proposal Networksopenaccess.thecvf.com/content_cvpr_2017/poster/225... · 2018. 1. 23. · Object Detection in Videos with Tubelet Proposal

Object Detection in Videos with Tubelet Proposal Networks

Proposals in Video Object Detection Tubelet Proposal Networks Qualitative Results

Framework

Encoder-decoder LSTM

YouTube-Objects DatasetExperimental Settings

Kai Kang1, Hongsheng Li1, Tong Xiao1, Wanli Ouyang1,4, Junjie Yan2, Xihui Liu3, Xiaogang Wang11The Chinese University of Hong Kong, 2SenseTime Group Limited

3Tsinghua University, 4The University of Sydney

Results on ImageNet VID Dataset

Frames

Per-frame staticproposals

Regression to GT Boxes(RegressBox)

Regression toGT Movement

(RegressMove, Ours)

t

y

x

Tubelet Proposal Network

Motion Prediction EncoderLSTM

DecoderLSTM

ClassLabel

TubeletFeatures

Tubelet Proposal Network Encoder-decoder LSTM

ClassificationCNN

Spatial Anchors t

y

x

TubeletCNN

Tubelet GenerationA

B

A A A A

B

B

B

B

4

2f

16

5f

W2

b2

W5

b5

Block Initialization

Iter 1Iter 2

Iter 3Iter 4

Iter 5

l

Parallel Generation

Class Label

Decoder LSTM

Encoder LSTM

Tubelet Features

Object localization on YouTube-Objects (YTO) dataset

Qualitative results on ImageNet VID validation set

Results on ImageNet VID validation set

Initialization of tubelet proposal networks

Carry Object Detection

Carry Object Detection

Object Detection in Videos With Tubelet Proposal Networksopenaccess.thecvf.com/content_cvpr_2017/papers/Kang_Object... · Object Detection in Videos with Tubelet Proposal Networks

Object Detection in Videos With Tubelet Proposal Networksopenaccess.thecvf.com/content_cvpr_2017/papers/Kang_Object... · Object Detection in Videos with Tubelet Proposal Networks

wangshy31@gmail.com luhc15@mails.tsinghua.edu.cn michael ... · wangshy31@gmail.com luhc15@mails.tsinghua.edu.cn michael@tsinghua.edu.cn Abstract Object detection in videos has drawn

[email protected] [email protected] michael ... · [email protected] [email protected] [email protected] Abstract Object detection in videos has drawn

Object Detection - University of Adelaidejaven/talk/ML17_main.pdf · Object detection is a type of classification problem. Object detection requires that we locate a specific object

Object Detection - University of Adelaidejaven/talk/ML17_main.pdf · Object detection is a type of classification problem. Object detection requires that we locate a specific object

Object Detection by Matching Longin Jan Latecki. Contour-based object detection Database shapes: …..

Object Detection by Matching Longin Jan Latecki. Contour-based object detection Database shapes: …..

ultrasonic object detection

ultrasonic object detection

Schneiderman, H. and Kanade, T. Object Detection Using the Statistics of Parts, Viola, P. and Jones Robust Real-time Object Detection,Object Detection.

Schneiderman, H. and Kanade, T. Object Detection Using the Statistics of Parts, Viola, P. and Jones Robust Real-time Object Detection,Object Detection.

Abandoned Object Detection in Video - University of …webdocs.cs.ualberta.ca/~asingh1/docs/Abandoned Object Detection in...Singh titled ―Abandoned object detection in video‖ be

Abandoned Object Detection in Video - University of …webdocs.cs.ualberta.ca/~asingh1/docs/Abandoned Object Detection in...Singh titled ―Abandoned object detection in video‖ be

Object Detection by Matching

Object Detection by Matching

Object Detection with Superquadrics

Object Detection with Superquadrics

T-CNN: Tubelets with Convolutional Neural Networks for Object … · 1 T-CNN: Tubelets with Convolutional Neural Networks for Object Detection from Videos Kai Kang*, Hongsheng Li*,

T-CNN: Tubelets with Convolutional Neural Networks for Object … · 1 T-CNN: Tubelets with Convolutional Neural Networks for Object Detection from Videos Kai Kang*, Hongsheng Li*,

Moving Object Detection and Tracking in Videos

Moving Object Detection and Tracking in Videos

Object of interest discovery in video sequences€¦ · images and videos. Traditionally, the analysis was done using object detection involving numerous human labors for labeling

Object of interest discovery in video sequences€¦ · images and videos. Traditionally, the analysis was done using object detection involving numerous human labors for labeling

Automatic identification of persons in TV seriesstiefel/... · videos are: automatic structuring or segmentation of videos, automatic speech recognition, object detection, concept

Automatic identification of persons in TV seriesstiefel/... · videos are: automatic structuring or segmentation of videos, automatic speech recognition, object detection, concept

Object Detection in Surveillance Video from Dense Trajectoriesmori/research/papers/zhai-mva15.pdf · Object Detection in Surveillance Video from Dense Trajectories ... Object detection

Object Detection in Surveillance Video from Dense Trajectoriesmori/research/papers/zhai-mva15.pdf · Object Detection in Surveillance Video from Dense Trajectories ... Object detection

A MultiPath Network for Object Detection arXiv:1604 ... · 2015 detection and segmentation challenges. 1 Introduction Object classification [19,28,30] and object detection [10,27,29]

A MultiPath Network for Object Detection arXiv:1604 ... · 2015 detection and segmentation challenges. 1 Introduction Object classification [19,28,30] and object detection [10,27,29]

Acceleration of Multi-Object Detection and Classification ...on-demand.gputechconf.com/gtc/...multi-object-detection-classification... · Acceleration of Multi-Object Detection and

Acceleration of Multi-Object Detection and Classification ...on-demand.gputechconf.com/gtc/...multi-object-detection-classification... · Acceleration of Multi-Object Detection and

Suspicious Object Detection and Tracking · Suspicious Object Detection and Tracking 7 1.3 SCOPE OF THE PROJECT The application provides efficient “Suspicious Object Detection”

Suspicious Object Detection and Tracking · Suspicious Object Detection and Tracking 7 1.3 SCOPE OF THE PROJECT The application provides efficient “Suspicious Object Detection”

Object Tracking Review - KU ITTC · ROBUST OBJECT TRACKING AND ADAPTIVE DETECTION ... • Object Detection • Object Tracking ... Varying Object velocity. Rapid Background Change

Object Tracking Review - KU ITTC · ROBUST OBJECT TRACKING AND ADAPTIVE DETECTION ... • Object Detection • Object Tracking ... Varying Object velocity. Rapid Background Change

Moving object detection

Moving object detection