Learning to Infer Implicit Surfaces without 3D Supervision

Authors: Shichen Liu, Shunsuke Saito, Weikai Chen, Hao Li

NeurIPS 2019 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results demonstrate the superiority of our method over state-of-the-art unsupervised 3D deep learning techniques, that are based on alternative shape representations, in terms of quantitative and qualitative measures. Comprehensive ablation studies also verify the efficacy of proposed probing-based sampling technique and the implicit geometric regularization.
Researcher Affiliation Collaboration Shichen Liu , , Shunsuke Saito , , Weikai Chen (B) , and Hao Li , , USC Institute for Creative Technologies University of Southern California Pinscreen {liushichen95, shunsuke.saito16, chenwk891}@gmail.com hao@hao-li.com
Pseudocode No The information is insufficient. The paper describes its methods but does not include any structured pseudocode or algorithm blocks.
Open Source Code No The information is insufficient. The paper does not contain any explicit statement about providing open-source code for the described methodology or a link to a code repository.
Open Datasets Yes We evaluate our method on Shape Net [10] dataset. We focus on 6 commonly used categories with complex topologies: plane, bench, table, car, chair and boat. We use the same train/validate/test split as in [4, 1, 2] and the rendered images (64 64 resolution) provided by [1] which consist of 24 views for each object.
Dataset Splits Yes We use the same train/validate/test split as in [4, 1, 2] and the rendered images (64 64 resolution) provided by [1] which consist of 24 views for each object.
Hardware Specification Yes We train the network using Adam optimizer with learning rate of 1 10 4 and batch size of 8 on a single 1080Ti GPU.
Software Dependencies No The information is insufficient. The paper mentions using a pre-trained ResNet18 and Adam optimizer but does not specify version numbers for any software dependencies like deep learning frameworks, Python, or CUDA.
Experiment Setup Yes Implementation details. We adopt a pre-trained Res Net18 as the encoder, which outputs a latent code of 128 dimensions. The decoder is realized using 6 fully-connected layers (output channels as 2048, 1024, 512, 256, 128 and 1 respectively) followed by a sigmoid activation function. We sample Np = 16, 000 anchor points in 3D space and Nr = 4096 rays for each view. The sampling bandwidth σ is set as 7 10 3. The radius τ of the supporting region is set as 3 10 2. For the regularizer, we set d = 3 10 2, λ = 1 10 2, and norm p = 0.8. We train the network using Adam optimizer with learning rate of 1 10 4 and batch size of 8 on a single 1080Ti GPU.