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; scores should be interpreted as estimates. Full accuracy metrics and methodology are described in [1]

ESEG: Event-Based Segmentation Boosted by Explicit Edge-Semantic Guidance

Authors: Yucheng Zhao, Gengyu Lyu, Ke Li, Zihao Wang, Hao Chen, Zhen Yang, Yongjian Deng

AAAI 2025 | Venue PDF | LLM Run Details

Reproducibility Variable Result LLM Response
Research Type Experimental Experimental results on DSEC and DDD17 datasets demonstrate the efficacy of the ESEG framework and its core designs.
Researcher Affiliation Academia 1College of Computer Science, Beijing University of Technology 2School of Computer Science and Engineering, Southeast University
Pseudocode No The paper describes the SELSAM algorithm and D2CAF module with equations and diagrams (e.g., Figure 3, Figure 4) but does not present structured pseudocode or algorithm blocks.
Open Source Code Yes Code https://github.com/cheesewawa/ESEG
Open Datasets Yes Experiments for this paper are conducted on two commonly used datasets for event-based semantic segmentation, DSEC-Semantic (Gehrig et al. 2021) and DDD17 (Binas et al. 2017).
Dataset Splits No The paper refers to using DSEC-Semantic and DDD17 datasets for training and evaluation but does not specify the exact training, validation, and test splits (e.g., percentages, sample counts, or specific predefined split references) in the main text. It mentions '11-class labels are used for training and evaluation' for DSEC-Semantic, but not the data splits.
Hardware Specification Yes All experiments are implemented using Pytorch on an RTX 3090.
Software Dependencies No The paper mentions implementing experiments using Pytorch but does not specify its version or other software dependencies with version numbers.
Experiment Setup Yes The Adam W optimizer and the Polynomial LR scheduler are used with the initial learning rate as 6e 5. The model was trained for 40 epochs with the batch size as 4. ... The learning rate scheduler is the same as DSEC but the initial learning rate is 1 10 3 for DDD17.