Deep Occlusion Reasoning for Multi-camera Multi-target Detection (2017-04-19T00:00:00.000000Z)

TL;DR

A new architecture is introduced that combines Convolutional Neural Nets and Conditional Random Fields to explicitly model ambiguities in people detection, and it is shown that it outperforms several state-of-the-art algorithms on challenging scenes.

Abstract

People detection in single 2D images has improved greatly in recent years. However, comparatively little of this progress has percolated into multi-camera multipeople tracking algorithms, whose performance still degrades severely when scenes become very crowded. In this work, we introduce a new architecture that combines Convolutional Neural Nets and Conditional Random Fields to explicitly model those ambiguities. One of its key ingredients are high-order CRF terms that model potential occlusions and give our approach its robustness even when many people are present. Our model is trained end-to-end and we show that it outperforms several state-of-the-art algorithms on challenging scenes.

Authors

P. Fua

9 papers

P. Baqué

3 papers

F. Fleuret

2 papers

TL;DR

Abstract

Authors

References32 items

The WILDTRACK Multi-Camera Person Dataset

Deep Multi-camera People Detection

Learning Arbitrary Potentials in CRFs with Gradient Descent

Social Scene Understanding: End-to-End Multi-person Action Localization and Collective Activity Recognition

Multi-modal Mean-Fields via Cardinality-Based Clamping

Multi-view People Tracking via Hierarchical Trajectory Composition

Theano: A Python framework for fast computation of mathematical expressions

How Far are We from Solving Pedestrian Detection?

Variational Inference: A Review for Statisticians

SSD: Single Shot MultiBox Detector

Deep Neural Decision Forests

Higher Order Potentials in End-to-End Trainable Conditional Random Fields

Principled Parallel Mean-Field Inference for Discrete Random Fields

Joint Training of Generic CNN-CRF Models with Stochastic Optimization

You Only Look Once: Unified, Real-Time Object Detection

Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks

Robust multiple cameras pedestrian detection with multi-view Bayesian network

Conditional Random Fields as Recurrent Neural Networks

Learning Graphical Model Parameters with Approximate Marginal Inference

Filter-Based Mean-Field Inference for Random Fields with Higher-Order Terms and Product Label-Spaces

Higher-Order Models in Computer Vision

Sparsity Driven People Localization with a Heterogeneous Network of Cameras

Neural conditional random fields

Framework for Performance Evaluation of Face, Text, and Vehicle Detection and Tracking in Video: Data, Metrics, and Protocol

Graphical Models, Exponential Families, and Variational Inference

Multicamera People Tracking with a Probabilistic Occupancy Map

Pattern Recognition and Machine Learning

Probability Product Kernels

Image analysis for video surveillance based on spatial regularization of a statistical model-based change detection

A Bayesian Computer Vision System for Modeling Human Interactions

A Tutorial on Energy-Based Learning

Ieee Transactions on Pattern Analysis and Machine Intelligence 1 Multiple Object Tracking Using K-shortest Paths Optimization

Field of Study

Journal Information

Name

Page

Venue Information

Name

Type

URL

Alternate Names