Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
A Boosted Multi-Task Model for Pedestrian Detection with Occlusion Handling
Authors: Chao Zhu, Yuxin Peng
AAAI 2015 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | To evaluate the proposed approach, we carry out experiments on the challenging Caltech pedestrian detection benchmark (Doll ar et al. 2012), and achieve state-of-the-art performances on the most popular Reasonable and three occlusion-specific test sets. |
| Researcher Affiliation | Academia | Chao Zhu and Yuxin Peng Institute of Computer Science and Technology, Peking University Beijing 100871, China EMAIL |
| Pseudocode | Yes | Algorithm 1: Multi-Task Decision Stumps Learning and Algorithm 2: Multi-Task ACF Model Learning |
| Open Source Code | No | The paper does not provide any explicit statement or link indicating that its source code is open or publicly available. |
| Open Datasets | Yes | The experiments are conducted on the Caltech pedestrian detection benchmark (Doll ar et al. 2012), which is by far the largest, most realistic and challenging pedestrian dataset. ... 1www.vision.caltech.edu/Image Datasets/Caltech Pedestrians/ |
| Dataset Splits | Yes | We follow a common training-testing protocol as in the literature: the pedestrian detector is trained on its training set (set00-set05), and the results are reported on its test set (set06-set10). |
| Hardware Specification | No | The paper does not provide any specific hardware details (e.g., CPU, GPU models, memory) used for running the experiments. |
| Software Dependencies | No | The paper mentions 'evaluation software (version 3.2.0) provided by Doll ar et al.' for evaluation, but does not list specific software dependencies with version numbers (e.g., programming languages, libraries, frameworks) used for implementing and running their proposed model. |
| Experiment Setup | Yes | The training parameters in the proposed approach are set as follows. 2048 weak classifiers are trained and combined to a strong classifier, and the nodes of the decision trees are constructed using a pool of 30,000 candidate regions from image samples. The multi-scale models are used to increase scale invariance. Two bootstrapping stages are applied with 5000 additional hard negatives each time. |