Consistent Structural Relation Learning for Zero-Shot Segmentation

Authors: Peike Li, Yunchao Wei, Yi Yang

NeurIPS 2020 | Conference PDF | Archive PDF | Plain Text | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental We conduct extensive experiments on Pascal-VOC and Pascal-Context benchmarks. The proposed CSRL outperforms existing state-of-the-art methods by a large margin, resulting in ~7-12% on Pascal-VOC and ~2-5% on Pascal-Context.
Researcher Affiliation Collaboration Peike Li1,2 , Yunchao Wei1, Yi Yang1 1Re LER Lab, Australian Artificial Intelligence Institute University of Technology Sydney 2Baidu Research peike.li@student.uts.edu.au, {yunchao.wei, yi.yang}@uts.edu.au
Pseudocode No The paper does not contain any structured pseudocode or algorithm blocks.
Open Source Code No The paper states: "We implemented our method both by the Pytorch platform and the Paddle Paddle platform, both achieving similar performance." However, it does not provide any link or explicit statement about the release of its source code.
Open Datasets Yes We conduct experiments on two datasets including Pascal-VOC [50] and Pascal Context [51].
Dataset Splits Yes Pascal-VOC focuses on object semantic segmentation scenario, which contains 10,582 training and 1,449 validation images from 20 classes. Pascal-Context targets on the scene parsing scenario, which comprises 4,998 training and 5,105 validation images from 59 classes.
Hardware Specification No The paper does not provide specific details about the hardware used for running the experiments (e.g., GPU models, CPU types).
Software Dependencies No The paper states: "We implemented our method both by the Pytorch platform and the Paddle Paddle platform..." However, it does not specify version numbers for these software platforms or any other dependencies.
Experiment Setup No The paper states: "More details of the network structure and parameter settings can be found in our supplementary materials." This indicates that specific experimental setup details like hyper-parameters are not in the main text.