SURGE: Surface Regularized Geometry Estimation from a Single Image
Authors: Peng Wang, Xiaohui Shen, Bryan Russell, Scott Cohen, Brian Price, Alan L. Yuille
NeurIPS 2016 | Conference PDF | Archive PDF | Plain Text | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We show that our regularization yields a 30% relative improvement in planar consistency on the NYU v2 dataset [24]. and We perform all our experiments on the NYU v2 dataset [24]. |
| Researcher Affiliation | Collaboration | 1University of California, Los Angeles 2Adobe Research 3Johns Hopkins University |
| Pseudocode | No | The paper does not contain structured pseudocode or clearly labeled algorithm blocks. |
| Open Source Code | No | The paper does not provide concrete access to source code (specific repository link, explicit code release statement, or code in supplementary materials) for the methodology described in this paper. |
| Open Datasets | Yes | We perform all our experiments on the NYU v2 dataset [24]. It contains 1449 images with size of 640 480, which is split to 795 training images and 654 testing images. and 1http://cs.nyu.edu/~silberman/datasets/nyu_depth_v2.html |
| Dataset Splits | No | It contains 1449 images with size of 640 480, which is split to 795 training images and 654 testing images. (No explicit validation set size or split methodology provided beyond mentioning its use). |
| 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 implementing algorithms based on 'Caffe [13]' but does not specify a version number for Caffe or any other software dependencies. |
| Experiment Setup | Yes | We empirically set σn = 0.1 for normals prediction and σd = 0.15 for depth prediction to produce reasonable confidence values. and For the surface bilateral filter in Eqn. (5), we set the parameters θα = 0.1, θβ = 50, θγ = 3, ω1 = 1, ω2 = 0.3, and set the λ = 2 in Eqn.(1) through a grid search over a validation set from [9]. |