Volume Rendering of Neural Implicit Surfaces

Authors: Lior Yariv, Jiatao Gu, Yoni Kasten, Yaron Lipman

NeurIPS 2021 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We evaluate our method on the challenging task of multiview 3D surface reconstruction. We use two datasets: DTU [12] and Blended MVS [37], both containing real objects with different materials that are captured from multiple views. In Section 4.1 we show qualitative and quantitative 3D surface reconstruction results of Vol SDF, comparing favorably to relevant baselines.
Researcher Affiliation Collaboration Lior Yariv1 Jiatao Gu2 Yoni Kasten1 Yaron Lipman1,2 1Weizmann Institute of Science 2Facebook AI Research
Pseudocode Yes Algorithm 1: Sampling algorithm. Input: error threshold ϵ > 0; β 1 Initialize T = T0 2 Initialize β+ such that BT ,β+ ϵ 3 while BT ,β > ϵ and not max_iter do 4 upsample T 5 if BT ,β+ < ϵ then 6 Find β (β, β+) so that BT ,β = ϵ 7 Update β+ β 10 Estimate b O using T and β+ 11 S get fresh m samples using ˆO 1 12 return S
Open Source Code No The paper does not include an explicit statement about releasing source code or provide a link to a code repository.
Open Datasets Yes We use two datasets: DTU [12] and Blended MVS [37], both containing real objects with different materials that are captured from multiple views.
Dataset Splits No The paper mentions training on a collection of images and evaluating on specific scans from the DTU and Blended MVS datasets, but it does not explicitly provide details about the training, validation, and test data splits (e.g., percentages, sample counts, or references to predefined splits within this paper).
Hardware Specification No The paper does not provide any specific details about the hardware used for running the experiments (e.g., GPU models, CPU types, or cloud computing instances).
Software Dependencies No The paper does not provide specific details about software dependencies, such as programming languages, libraries, or frameworks, along with their version numbers.
Experiment Setup Yes We train with batches of size 1024 pixels p. λ is a hyper-parameter set to 0.1 throughout the the experiments.