robustness-analysis

Unified Multilingual Robustness Evaluation Toolkit for Natural Language Processing

A lightweight Python package for setting up robustness experiments and to compute robustness distributions.

NAACL 2022 paper on Analyzing Modality Robustness in Multimodal Sentiment Analysis

Official Code for "Can These Views Be One Scene?"

Potential of 2D Priors for Improving Robustness of Ill-Posed 3D Reconstruction

[Elsevier Image and Vision Computing] How robust are discriminatively trained zero-shot learning models?

Analyzing and Improving the Robustness of Tabular Classifiers using Counterfactual Explanations

Official repository for the paper: "On Adversarial Training without Perturbing all Examples", Accepted at ICLR 2024

To verify and analyze classification properties of the neural networks when a small perturbation is applied to the image from MNIST, CIFAR-10 (and both with prepared Blured-Image) dataset.

𝒮𝒟-2 · System Deviation Diagnosis — a robustness diagnosis framework for end-to-end (E2E) autonomous driving. Decomposes the driving pipeline (vision → semantic → planning → control → outcome), measures stage-wise deviation between clean and stress CARLA runs, and localizes where robustness first collapses (InterFuser, TransFuser).

Analyzing semantic segmentation robustness in adverse-weather driving scenes

Built on JSBSim, a high-fidelity swarm drone beyond-visual-range (BVR) air combat simulation platform features 6-DOF flight dynamics, credible AIM-120C-5 missile physics, non-ideal perception with distance-scaled noise, absolute-state Kalman filter, parallel sandboxed strategies, and Tacview-compatible ACMI replay for validating strategy robustness

An advanced DCT-based steganography framework analyzing embedding behavior, capacity limits and resistance to JPEG compression and noise attacks.

Investigating the robustness of deep reinforcement learning for unmanned surface vehicle interception under target observation loss, combining PPO, policy distillation, Kalman prediction, and paired simulation experiments.

Linear Kalman filtering for simulated 2D GPS/IMU state estimation, with dropout, noise, and acceleration-bias sensitivity experiments.

Demonstration of robustness analysis for (multistage) adaptive optimization problems

B.Sc. Thesis Project: A research-driven framework for protecting intellectual property in relational databases using digital watermarking, focusing on data integrity and resistance against tuple-based attacks.

Revised implementation for the paper “Blind confusion of classification networks”, evaluating image classification models under common and structured corruptions.