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 Coakley et alK. L. Coakley, T. Snelleman, H. Hoos, and O. E. Gundersen, "The embrace of open science: An analysis of a decade of AI research and 56 800 conference papers," Under Review, 2026..
Distil-E2D: Distilling Image-to-Depth Priors for Event-Based Monocular Depth Estimation
Authors: Jie Long Lee, Gim Hee Lee
NeurIPS 2025 | Venue PDF | LLM Run Details
| Reproducibility Variable | Result | LLM Response |
|---|---|---|
| Research Type | Experimental | Experiments on benchmark datasets show that Distil-E2D achieves state-of-the-art performance in event-based monocular depth estimation across both event-only and event+APS settings. |
| Researcher Affiliation | Academia | Jie Long Lee Gim Hee Lee Department of Computer Science, National University of Singapore EMAIL, EMAIL |
| Pseudocode | No | The paper describes the methodology using textual explanations and mathematical equations (e.g., Eq. 1-15) and details the network architecture in Appendix A.2, but does not provide any explicitly labeled pseudocode or algorithm blocks. |
| Open Source Code | Yes | Code available at the project website1. 1https://github.com/leejielong/Distil-E2D |
| Open Datasets | Yes | Experiments on benchmark datasets...MVSEC Dataset [2], DSEC Dataset [3]...A.5 License and Credit: MVSEC Dataset: Creative Commons BY 4.0 License, DSEC Dataset: Creative Commons BY-SA 4.0 License |
| Dataset Splits | Yes | MVSEC. This dataset [2] is a standard benchmark for event-based depth estimation...Following prior work, we use the outdoor_day_2 sequence for training and evaluate on the remaining four sequences. 2) DSEC...We use 28 sequences for training and 13 for testing. |
| Hardware Specification | Yes | All experiments were performed on a single NVIDIA RTX A6000 GPU with 48 GB of memory. |
| Software Dependencies | Yes | Distil-E2D is implemented using Py Torch 2.0 and trained with the Adam W optimizer |
| Experiment Setup | Yes | Distil-E2D is implemented using Py Torch 2.0 and trained with the Adam W optimizer at a learning rate of 1 10 4. A One Cycle LR scheduler dynamically adjusts the learning rate. Training is conducted over 250 epochs with full precision and a batch size of 16, using sequences of 10 frames corresponding to a 500ms event window. |