Object Localization based on Structural SVM using Privileged Information

Authors: Jan Feyereisl, Suha Kwak, Jeany Son, Bohyung Han

NeurIPS 2014 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We apply the proposed algorithm to the Caltech-UCSD Birds 200-2011 dataset, and obtain encouraging results suggesting further investigation into the benefit of privileged information in structured prediction. We evaluate our method by learning to localize birds in the Caltech-UCSD Birds 200-2011 (CUB-2011) dataset and exploiting attributes and segmentation masks as privileged information in addition to standard visual features.
Researcher Affiliation Academia Jan Feyereisl, Suha Kwak , Jeany Son, Bohyung Han Dept. of Computer Science and Engineering, POSTECH, Pohang, Korea thefillm@gmail.com, {mercury3,jeany,bhhan}@postech.ac.kr Current affiliation: INRIA WILLOW Project, Paris, France; e-mail: suha.kwak@inria.fr
Pseudocode Yes Algorithm 1 Cutting plane method for solving Eq. (6)
Open Source Code No The paper does not explicitly provide a link to its source code or state that it will be made publicly available.
Open Datasets Yes Empirical evaluation of our method is performed on the Caltech-UCSD Birds 2011 (CUB-2011) [5] fine-grained categorization dataset.
Dataset Splits Yes As a validation set, 500 training images chosen at random from categories other than the ones used for training are used.
Hardware Specification No The paper does not provide specific details about the hardware used for running its experiments.
Software Dependencies No The paper mentions using 'bag-of-visual-words model based on Speeded Up Robust Features (SURF) [26]' but does not provide specific version numbers for any software libraries, frameworks, or programming languages used for implementation.
Experiment Setup Yes In all experiments we tune the hyperparameters C, λ and ρ on a 4 4 4 space spanning values [2 8, ..., 25].