8000 FarinaMatteo (Matteo Farina) · GitHub
[go: up one dir, main page]
More Web Proxy on the site http://driver.im/
Skip to content
View FarinaMatteo's full-sized avatar

Highlights

  • Pro

Block or report FarinaMatteo

Block user

Prevent this user from interacting with your repositories and sending you notifications. Learn more about blocking users.

You must be logged in to block users.

Please don't include any personal information such as legal names or email addresses. Maximum 100 characters, markdown supported. This note will be visible to only you.
Report abuse

Contact GitHub support about this user’s behavior. Learn more about reporting abuse.

Report abuse
FarinaMatteo/README.md

Hello there! 👋

🔭 Currently.

⌛️ Previously.

  • Research Intern with Elisa Ricci and Federica Arrigoni (University of Trento), Vladislav Golyanik (Max Planck Institute for Informatics), Luca Magri (Politecnico di Milano), working on Quantum Computer Vision.
  • BS/MS Student @ the University of Trento.

💡 My work revolves around Large Multimodal Models, especially their zero-shot, few-shot, and test-time generalization. I approach this from two angles: ☝️ pretrained models' behavior—understanding where public models fail and how to improve them and ✌️ data-centric pretraining—exploring how smarter pretraining and better-curated data create stronger learners.

Previously, I worked on model compression sponsored by a Cisco Research Grant.

📫 Feel free to reach out here, I'm always open to collabs!

Pinned Loading

  1. multiflow multiflow Public

    [CVPR '24] Official implementation of the paper "Multiflow: Shifting Towards Task-Agnostic Vision-Language Pruning".

    Python 23 2

  2. zero zero Public

    [NeurIPS '24] Frustratingly easy Test-Time Adaptation of VLMs!!

    Python 47 3

  3. qmmf qmmf Public

    [CVPR '23 Highlight] Official repository for the paper "Quantum Multi-Model Fitting".

    Python 11 4

  4. rethinking_fewshot_vlms rethinking_fewshot_vlms Public

    [CVPR '25] Official implementation of the paper "Rethinking Few-Shot Adaptation of Vision-Language Models in Two Stages", accepted at (and to appear in) CVPR 2025.

    Python 15 1

0