Multimodal Residual Learning for Visual QA

Authors: Jin-Hwa Kim, Sang-Woo Lee, Donghyun Kwak, Min-Oh Heo, Jeonghee Kim, Jung-Woo Ha, Byoung-Tak Zhang

NeurIPS 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We achieve the state-of-the-art results on the Visual QA dataset for both Open-Ended and Multiple-Choice tasks.
Researcher Affiliation Collaboration Jin-Hwa Kim Sang-Woo Lee Donghyun Kwak Min-Oh Heo Seoul National University {jhkim,slee,dhkwak,moheo}@bi.snu.ac.kr Jeonghee Kim Jung-Woo Ha Naver Labs, Naver Corp. {jeonghee.kim,jungwoo.ha}@navercorp.com Byoung-Tak Zhang Seoul National University & Surromind Robotics btzhang@bi.snu.ac.kr
Pseudocode No The paper does not contain structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide concrete access to source code for the described methodology (Multimodal Residual Networks) via a specific repository link, explicit code release statement, or code in supplementary materials.
Open Datasets Yes We choose the Visual QA (VQA) dataset [1] for the evaluation of our models. [1] Aishwarya Agrawal, Jiasen Lu, Stanislaw Antol, Margaret Mitchell, C. Lawrence Zitnick, Dhruv Batra, and Devi Parikh. VQA: Visual Question Answering. In International Conference on Computer Vision, 2015.
Dataset Splits Yes The images come from the MS-COCO dataset, 123,287 of them for training and validation, and 81,434 for test. All validation is performed on the test-dev split.
Hardware Specification No The paper does not provide specific hardware details (exact GPU/CPU models, processor types with speeds, memory amounts, or detailed computer specifications) used for running its experiments.
Software Dependencies No The paper mentions 'Torch framework and rnn package [13]' and 'Python Natural Language Toolkit (nltk) [3]' but does not provide specific version numbers for these or other key software components.
Experiment Setup Yes The common embedding size of the joint representation is 1,200. The learnable parameters are initialized using a uniform distribution from 0.08 to 0.08 except for the pretrained models. The batch size is 200, and the number of iterations is fixed to 250k. The RMSProp [26] is used for optimization, and dropouts [7, 5] are used for regularization.