Multimedia and Human Understanding Group

university

http://mhug.disi.unitn.it/#/research

Activity Feed Request to join this org

AI & ML interests

Multimedia and Human Understanding Group, DISI @ UNITN, Trento, Italy

Papers

A Stationary (and Therefore Compatible) Representation is All You Need

Finetune Like You Pretrain: Boosting Zero-shot Adversarial Robustness in Vision-language Models

View all Papers

MHUGLab 's papers 8

Submitted by

Niccolò Biondi

A Stationary (and Therefore Compatible) Representation is All You Need

MHUGLab

Multimedia and Human Understanding Group

3

Finetune Like You Pretrain: Boosting Zero-shot Adversarial Robustness in Vision-language Models

MHUGLab

Multimedia and Human Understanding Group

Submitted by

Samuele Angheben

Specificity-aware reinforcement learning for fine-grained open-world classification

MHUGLab

Multimedia and Human Understanding Group

Submitted by

Marco Garosi

Large Multimodal Models as General In-Context Classifiers

MHUGLab

Multimedia and Human Understanding Group

Submitted by

Francesco Laiti

How to Take a Memorable Picture? Empowering Users with Actionable Feedback

MHUGLab

Multimedia and Human Understanding Group

Submitted by

Markus

Loomis Painter: Reconstructing the Painting Process

MHUGLab

Multimedia and Human Understanding Group

Submitted by

Alessandro Conti

On Large Multimodal Models as Open-World Image Classifiers

MHUGLab

Multimedia and Human Understanding Group

CLIP is Strong Enough to Fight Back: Test-time Counterattacks towards Zero-shot Adversarial Robustness of CLIP

MHUGLab

Multimedia and Human Understanding Group