Combining Models from Multiple Sources for RGB-D Scene Recognition

Authors: Xinhang Song, Shuqiang Jiang, Luis Herranz

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

Reproducibility Variable Result LLM Response
Research Type Experimental We evaluate our approach in two datasets: NYU Depth Dataset version 2 (NYUD2) [Silberman et al., 2012] and SUN RGB-D [Song et al., 2015]. The comparisons are reported in Table 3, where the proposed multi-source model for depth modality combines two models, improving the performance up to 40.1%.
Researcher Affiliation Academia 1Key Lab of Intel. Inf. Proc., Inst. of Comput. Tech., Chinese Academy of Sciences, Beijing, 100190, China 2University of Chinese Academy of Sciences, Beijing, 100049, China 3Computer Vision Center, 08193 Bellaterra, Barcelona, Spain
Pseudocode No The paper does not contain any structured pseudocode or algorithm blocks.
Open Source Code No The paper does not provide any concrete access to source code for the methodology described.
Open Datasets Yes We evaluate our approach in two datasets: NYU Depth Dataset version 2 (NYUD2) [Silberman et al., 2012] and SUN RGB-D [Song et al., 2015].
Dataset Splits Yes Following the publicly available split in [Song et al., 2015; Wang et al., 2016], the 19 most common categories are selected, consisting of 4845/4659 images for training/test, where the training set consists of a split of 2393/2452 images for training/validation.
Hardware Specification No The paper does not provide specific details about the hardware (e.g., GPU/CPU models, memory) used for running its experiments.
Software Dependencies No The paper mentions using 'SVM [Fan et al., 2008] classifiers' but does not provide specific version numbers for any software libraries or dependencies used in the experiments.
Experiment Setup Yes We empirically set λ = 0.1 and consider two settings.