FANTrack: 3D Multi-Object Tracking with Feature Association Network [on] [at] [la] [FANTrack]

Submitted on 6 Jun. 2019 05:42 by
Venkateshwaran Balasubramanian (WISE Lab - University of Waterloo)

Running time:0.04 s
Environment:8 cores @ >3.5 Ghz (Python)

Method Description:
Our approach uses a learning-based data association
framework in a tracking-by-detection setting. We
use a Siamese based feature extractor to generate
local similarity maps and feed them as input to a
CNN to solve the data association problem.
Detection Threshold = 0.28
Latex Bibtex:
title={FANTrack: 3D Multi-Object Tracking with
Feature Association Network},
author={Erkan Baser and Venkateshwaran
Balasubramanian and Prarthana Bhattacharyya and
Krzysztof Czarnecki},

Detailed Results

From all 29 test sequences, our benchmark computes the HOTA tracking metrics (HOTA, DetA, AssA, DetRe, DetPr, AssRe, AssPr, LocA) [1] as well as the CLEARMOT, MT/PT/ML, identity switches, and fragmentation [2,3] metrics. The tables below show all of these metrics.

Benchmark HOTA DetA AssA DetRe DetPr AssRe AssPr LocA
CAR 60.85 % 64.36 % 58.69 % 69.17 % 80.82 % 60.78 % 88.94 % 84.72 %

Benchmark TP FP FN
CAR 28130 6262 1305

CAR 75.84 % 82.46 % 78.00 % 743 61.49 %

Benchmark MT rate PT rate ML rate FRAG
CAR 62.77 % 28.46 % 8.77 % 701

Benchmark # Dets # Tracks
CAR 29435 1582

This table as LaTeX

This figure as: png pdf

[1] J. Luiten, A. Os̆ep, P. Dendorfer, P. Torr, A. Geiger, L. Leal-Taixé, B. Leibe: HOTA: A Higher Order Metric for Evaluating Multi-object Tracking. IJCV 2020.
[2] K. Bernardin, R. Stiefelhagen: Evaluating Multiple Object Tracking Performance: The CLEAR MOT Metrics. JIVP 2008.
[3] Y. Li, C. Huang, R. Nevatia: Learning to associate: HybridBoosted multi-target tracker for crowded scene. CVPR 2009.

eXTReMe Tracker