Boosted Zero-Shot Learning with Semantic Correlation Regularization

Authors: Te Pi, Xi Li, Zhongfei (Mark) Zhang

IJCAI 2017 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental The experiments on two ZSL datasets show the superiority of BZ-SCR over the state-of-the-arts.
Researcher Affiliation Collaboration Te Pi1, Xi Li1,2 , Zhongfei (Mark) Zhang1 1Zhejiang University, Hangzhou, China 2Alibaba-Zhejiang University Joint Institute of Frontier Technologies, Hangzhou, China
Pseudocode Yes Algorithm 1: Boosted Zero-shot classification with Semantic Correlation Regularization (BZ-SCR)
Open Source Code No The paper does not provide an explicit statement or link for the open-source code for the methodology described. It refers to external datasets but not its own code.
Open Datasets Yes We evaluate the performance of BZ-SCR on the two classic ZSL image datasets, Animal With Attributes1 (AWA) and Caltech-UCSD-Birds-2002 (CUB200). 1http://attributes.kyb.tuebingen.mpg.de/ 2http://www.vision.caltech.edu/visipedia/CUB-200-2011.html
Dataset Splits Yes For the class split, we adopt the default split (40/10 for train+val/test) for AWA and the same split as [Akata et al., 2013] (150/50 for train+val/test) for CUB200.
Hardware Specification No The paper does not explicitly mention any specific hardware used for running the experiments (e.g., GPU models, CPU types, or cloud configurations).
Software Dependencies No The paper does not provide specific version numbers for any key software components or libraries used in the implementation.
Experiment Setup Yes The shown results of BZ-SCR are achieved with ν/N = 10 4, β/N = 0.4 for AWA, and ν/N {0.025, 0.05}, β/N {0.1, 0.2} for CUB200. We implement a grid search for the tuning of the hyperparameters ν, β, and report the best performances.