Similarity Mapping with Enhanced Siamese Network for Multi-Object Tracking

Abstract

Multi-object tracking has recently become an important area of computer vision, especially for Advanced Driver Assistance Systems (ADAS). Despite growing attention, achieving high performance tracking is still challenging, with state-of-the- art systems resulting in high complexity with a large number of hyper parameters. In this paper, we focus on reducing overall system complexity and the number hyper parameters that need to be tuned to a specific environment. We introduce a novel tracking system based on similarity mapping by Enhanced Siamese Neural Network (ESNN), which accounts for both appearance and geometric information, and is trainable end-to-end. Our system achieves competitive performance in both speed and accuracy on MOT16 challenge, compared to known state-of-the-art methods.

1) accepted as a poster presentation at WiML (Women in Machine Learning) workshop 2016, colocated with NIPS 2016 in Barcelona, Spain, 2) accepted as a poster presentation at MLITS (Machine Learning for Intelligent Transportation Systems) Workshop held in conjunction with the NIPS 2016 in Barcelona, Spain

References in corpus (2)

• MOT16: A Benchmark for Multi-Object Tracking
• First Step toward Model-Free, Anonymous Object Tracking with Recurrent Neural Networks

Cited by in corpus (6)

• Deep Learning in Video Multi-Object Tracking: A Survey
• Simple Unsupervised Multi-Object Tracking
• GCNNMatch: Graph Convolutional Neural Networks for Multi-Object Tracking via Sinkhorn Normalization
• Simple online and real-time tracking with occlusion handling
• Enhancing the Association in Multi-Object Tracking via Neighbor Graph
• Extending Multi-Object Tracking systems to better exploit appearance and 3D information