Notice: The reproducibility variables underlying each score are classified using an automated LLM-based pipeline, validated against a manually labeled dataset. LLM-based classification introduces uncertainty and potential bias; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1].
Continuous Object Representation Networks: Novel View Synthesis without Target View Supervision
Authors: Nicolai Hani, Selim Engin, Jun-Jee Chao, Volkan Isler
NeurIPS 2020 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | We evaluate CORN on the task of novel view synthesis for several object classes and on potential applications, namely single-view 3D reconstruction and out of distribution view synthesis on real data. |
| Researcher Affiliation | Academia | Nicolai Häni Selim Engin Jun-Jee Chao Volkan Isler EMAIL University of Minnesota |
| Pseudocode | No | The paper describes its methods through textual explanations and mathematical equations, but it does not include any explicitly labeled pseudocode or algorithm blocks. |
| Open Source Code | Yes | For up-to-date information, data, and code, please see our project page 1. 1Project page: nicolaihaeni.github.io/corn/ |
| Open Datasets | Yes | For novel view synthesis, we follow established evaluation protocols [35, 49] and evaluate on the car and chair classes of Shape Net v2.0 [4]. |
| Dataset Splits | No | Of the 108 rendered images for each object, we select only two images at random per object for CORN training. We evaluate the performance of novel view synthesis on 20,000 randomly generated test pairs of objects in the held-out test dataset. The paper mentions training and test sets but does not specify a validation set split. |
| Hardware Specification | No | The authors acknowledge the Minnesota Supercomputing Institute (MSI) at the University of Minnesota for providing resources that contributed to the research results reported within this paper. This statement does not provide specific hardware details like GPU/CPU models or memory. |
| Software Dependencies | No | The paper mentions using ResNet-18 and UNet architectures and states that "Hyper-parameters, full network architectures for CORNs, and all baseline descriptions can be found in the supplementary material." However, it does not explicitly provide specific software names with version numbers in the main text. |
| Experiment Setup | No | The paper states "Hyper-parameters, full network architectures for CORNs, and all baseline descriptions can be found in the supplementary material." This indicates that experimental setup details are deferred to supplementary material and are not present in the main text. |