Robust MIL-Based Feature Template Learning for Object Tracking
Authors: Xiangyuan Lan, Pong C. Yuen, Rama Chellappa
AAAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experiments on challenging video sequences show that the proposed tracker performs favourably against several state-of-the-art trackers. |
| Researcher Affiliation | Academia | Department of Computer Science, Hong Kong Baptist University, Hong Kong SAR, China Center for Automation Research and ECE Department, University of Maryland, College Park, MD, USA |
| Pseudocode | Yes | Algorithm 1: Overall Optimization Procedure for (17) Algorithm 2: Procedure for Updating Feature Templates D in (17) Algorithm 3: Procedure for Updating Sparse Coefficients xij in (17) |
| Open Source Code | No | The paper states that source codes for compared trackers were used ('The source codes provided by the authors are used'), but does not provide concrete access to the source code for the proposed methodology. |
| Open Datasets | Yes | Twenty publicly available image sequences, which cover various kinds of challenging scenarios, e.g. occlusion, abrupt illumination changes, large pose variations, etc., are used for evaluation. |
| Dataset Splits | No | The paper describes how samples are generated and used for online model updating ('8 potentially positive samples and 50 samples in the first frame compose the positive bag for training'), but does not provide specific train/validation/test dataset splits for the overall evaluation dataset. |
| Hardware Specification | No | The paper does not provide specific hardware details (e.g., GPU/CPU models, memory) used for running its experiments. |
| Software Dependencies | No | The paper does not provide specific ancillary software details with version numbers (e.g., libraries, frameworks) needed to replicate the experiment. |
| Experiment Setup | Yes | We empirically set β in (8) to be 0.01, λ in (8) to be 0.01, λ1 in (19) to be 0.01, b in (8) to be 1, c in (8) to be 1, the number of the learned feature templates to be 20. |