Complex Event Detection via Event Oriented Dictionary Learning

Authors: Yan Yan, Yi Yang, Haoquan Shen, Deyu Meng, Gaowen Liu, Alex Hauptmann, Nicu Sebe

AAAI 2015 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Extensive experimental results on TRECVID Multimedia Event Detection (MED) dataset demonstrate the efficacy of our proposed method.
Researcher Affiliation Academia 1University of Trento, Italy 2University of Technology Sydney, Australia 3Xi an Jiao Tong University, China 4Carnegie Mellon University, USA
Pseudocode Yes Algorithm 1: Supervised Multi-task Dictionary Learning.
Open Source Code No The paper does not provide a link to open-source code for the described methodology or state that code is released.
Open Datasets Yes TRECVID MED10 (P001-P003) and MED11 (E001-E015) datasets are used in our experiments. The datasets consist of 9746 videos from 18 events of interest... TRECVID Semantic Indexing Task (SIN) contains annotation for 346 semantic concepts on 400,000 keyframes from web videos. 1http://www-nlpir.nist.gov/projects/tv2013/tv2013.html#sin 2http://www.nist.gov/itl/iad/mig/med12.cfm
Dataset Splits No There are 3104 videos used for training and 6642 videos used for testing in our experiments. No explicit validation split information is provided.
Hardware Specification No The paper does not provide specific hardware details such as GPU or CPU models used for running the experiments.
Software Dependencies No The paper mentions features (SIFT, CSIFT, MOSIFT) and algorithms (FISTA, Elastic-Net) but does not list specific software packages or libraries with version numbers required for reproduction.
Experiment Setup Yes We set the regularization parameters in the range of {0.01, 0.1, 1, 10, 100}. The subspace dimensionality s is set by searching the grid from {200, 400, 600}. For the experiments in the paper, we try three different dictionary sizes from {768, 1024, 1280}.