This repository is the official implementation of On the Black-box Explainability of Object Detection Models for Safe and Trustworthy Industrial Applications, where algorithms like D-RISE and D-MFPP specifically designed for Object Detection models can be... read more of On the Black-box Explainability of Object Detection Models for Safe and Trustworthy Industrial Applications

Understanding Face Attribute Classifiers through Summary Model Explanations. read more of FaceX

Comprehensive AI fairness exploration. read more of FairBench

Template to support the elicitation and prioritisation of trustworthiness requirements for a use case of the ULTIMATE project. read more of AI Trustworthiness User Requirements Template

A framework for evaluating LLM on iterative game content refinement read more of LLMAKER - Consistent Game Content Creation via LLMs

The repository contains the python code for the manifold mixing model soup algorithm as proposed in the paper at https://arxiv.org/abs/2309.08610. read more of Manifold mixing model soups for better out-of-distribution accuracy

Method that provides privacy-related feedback for photographic user profiles read more of LERVUP - LEarning to Rate Visual User Profiles

AdvisIL recommends appropriate algorithm-backbone architecture combinations for a user-provided class-incremental learning scenario. read more of AdvisIL - A Class-Incremental Learning Advisor

Transfer-learning-based method addressing exemplar-free class-incremental learning read more of Feature Translation for Exemplar-Free Class-Incremental Learning

FasterAI is an open-source library dedicated to optimizing AI models through advanced compression techniques like pruning, quantization, and knowledge distillation, ensuring they remain lightweight and efficient without sacrificing accuracy. It supports a... read more of FasterAI