alexklwong/stereoscopic-universal-perturbations
PyTorch Implementation of Stereoscopic Universal Perturbations across Different Architectures and Datasets (CVPR 2022)
This project helps researchers and engineers understand and mitigate a security vulnerability in deep learning systems used for stereo vision. It provides a way to generate specific, small, and visually unnoticeable modifications to stereo images (input) that can cause depth-sensing neural networks to make significant errors (output). This is useful for anyone working with depth perception in robotics, autonomous vehicles, or 3D mapping who needs to assess and improve the robustness of their deep learning models against subtle attacks.
No commits in the last 6 months.
Use this if you are developing or deploying deep stereo vision systems and need to evaluate their resilience to adversarial attacks or improve their robustness.
Not ideal if you are looking for a general image processing tool unrelated to deep learning or security vulnerabilities in stereo vision.
Stars
21
Forks
1
Language
Python
License
—
Category
Last pushed
Oct 13, 2022
Commits (30d)
0
Get this data via API
curl "https://pt-edge.onrender.com/api/v1/quality/diffusion/alexklwong/stereoscopic-universal-perturbations"
Open to everyone — 100 requests/day, no key needed. Get a free key for 1,000/day.
Higher-rated alternatives
anchen1011/toflow
TOFlow: Video Enhancement with Task-Oriented Flow
NVlabs/nvdiffrec
Official code for the CVPR 2022 (oral) paper "Extracting Triangular 3D Models, Materials, and...
RuojinCai/doppelgangers
Doppelgangers: Learning to Disambiguate Images of Similar Structures
cvg/GlueStick
Joint Deep Matcher for Points and Lines 🖼️💥🖼️ (ICCV 2023)
microsoft/SpareNet
Style-based Point Generator with Adversarial Rendering for Point Cloud Completion (CVPR 2021)