A Stochastic Image Grammar for Fine-Grained 3D Scene Reconstruction
Authors: Xiaobai Liu, Yadong Mu, Liang Lin
IJCAI 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Results with comparisons on public image dataset show that our method clearly outperforms the alternate state-of-the-art single-view reconstruction methods. |
| Researcher Affiliation | Academia | 1Department of Computer Science, San Diego State University, San Diego, 92182, CA, USA 2 Institute of Computer Science and Technology, Peking University, Beijing, 100871, China 3 School of Data and Computer Science, Sun Yat-Sen University, Guangzhou, 510006, China |
| Pseudocode | Yes | Algorithm 1 Inference . |
| Open Source Code | No | The paper does not provide an explicit statement about releasing source code or a link to a code repository for the methodology described. |
| Open Datasets | Yes | We use the three datasets [Liu et al., 2014]: CMU dataset, LMW-A, and LMW-B. The CMU dataset was originally collected by Hoiem et al. [Hoiem et al., 2008]... We further collect 950 images from different sources, i.e. LMW-C, and manually annotate VPs, region labels and surface normal orientations . These images are selected from the PASCAL VOC [Everingham et al., 2015] and Labelme projects [Russell et al., 2007]. |
| Dataset Splits | No | The paper mentions training and testing splits (e.g., 'We used 50 images for training and the rest for testing') but does not explicitly provide details about a separate validation split or its size/percentage. |
| Hardware Specification | Yes | For each image, the average processing time is 50 seconds on a Workstation(i7@3.6GHZ with 16GB memory). |
| Software Dependencies | No | The paper mentions 'Matlab optimization toolbox' but does not provide specific version numbers for any software dependencies. |
| Experiment Setup | Yes | To make proposals for R1 through R5, we set the spatial distance between two children nodes (i.e. edge or planar region) to be less than 40 pixels, and the orientation angles, if applicable, to be less than 10 degrees. ... The maximal iterations of HMC algorithm is fixed to 2000. |