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

By Rangachar Kasturi, Dmitry Goldgof, P. Soundararajan, V. Manohar, John S. Garofolo, Rachel Bowers, Matthew Boonstra, V. Korzhova, Jing Zhang. IEEE Transactions on Pattern Analysis and Machine Intelligence, 2009

562 citations recorded by the source corpus.

Subfields: Computer Vision and Pattern Recognition

View Paper: https://www.semanticscholar.org/paper/b0246b2edca38d7def294cb55faccc70c55c8a69

How researchers see this paper

The researchers whose trusted networks rank it highest, strongest advocate first. Every researcher gets a different order.

In Computer Vision and Pattern Recognition

Citations in the corpus

Looking at Vehicles on the Road: A Survey of Vision-Based Vehicle Detection, Tracking, and Behavior Analysis (2013, background): “(a) Overlap criterion used for labeling detections as true or false positives in the image plane [196].”
Humans in 4D: Reconstructing and Tracking Humans with Transformers (2023, methodology): “, SMPLify [7] and variants [22, 38, 56, 66, 72, 85], for this analysis we will focus on approaches that directly regress the body shape from a single image input.”
Visual Tracking: an Experimental Survey (2013, unknown): “The VACE metrics [39], [31] define the average tracking accuracy, ATA, slightly different from OTP : ATA = 1 Nframes ∑ i |T i ∩GT i| |T i ∪GT i| .”
The Visual Object Tracking VOT2015 Challenge Results (2015, background): “FERET [56] and [30], and the recent long-term tracking and detection of general targets6 to list but a few.”
Multiple Object Tracking Using K-Shortest Paths Optimization (2011, unknown): “These metrics have become standard for evaluation of detection and tracking algorithms, and we refer the interested reader to [41], [38] for more detailed descriptions and motivations.”
Data-driven Sequential Monte Carlo in Probabilistic Programming (2015, methodology): “To measure the performance, we use and report commonly used performance metrics: the sequence frame detection accuracy (SFDA) for object detection and the average tracking accuracy (ATA) for tracking [5].”
A Protocol for Evaluating Video Trackers Under Real-World Conditions (2013, background): “Moreover, because of the complexity of the evaluation task, many performance criteria contain multiple measures [4], [6], [7], which are difficult to combine in order to rank various algorithms.”
The Visual Object Tracking VOT2017 Challenge Results (2017, background): “FERET [50] and [28]), long-term tracking 6 and multiple target tracking [35, 61]7.”
A Novel Performance Evaluation Methodology for Single-Target Trackers (2015, background): “Visual tracking is a rapidly evolving field that has been increasingly attracting attention of the vision community.”
Understanding Transit Scenes: A Survey on Human Behavior-Recognition Algorithms (2010, background): “interest to the research surveillance community [9]–[17], [19].”
Learning a Neural Solver for Multiple Object Tracking (2019, background): “Both exact and greedy rounding show almost equal performance, with greedy rounding having slightly lower IDF1 (0.2 percentage points), equal MOTA and equal speed.”
The Visual Object Tracking VOT2014 Challenge Results (2014, unknown): “FERET [50] and [31], and the recent long-term tracking and detection of general targets(30) to list but a few.”
Unsupervised Detection and Tracking of Arbitrary Objects with Dependent Dirichlet Process Mixtures (2012, methodology): “The metrics presented in [37] and used in [22, 59, 40] have become well established for evaluating the performance of object detection and tracking in videos and have been adopted by the Video Analysis and Content Extraction (VACE) program and the Classification of Events, Activities, and Relationsh”
The Visual Object Tracking VOT2016 Challenge Results (2016, background): “FERET [12,13]), long-term tracking(5) and the multiple target tracking [14,15](6).”