vision-transformer

Adaptive Token Sampling for Efficient Vision Transformers (ECCV 2022 Oral Presentation)

🥶Vilio: State-of-the-art VL models in PyTorch & PaddlePaddle

Multi-modal AI-generated content detection: image, video, and audio. Benchmarks, training code (DINOv2, DINOv3, ReStraV, BreathNet), and evaluation pipeline for real vs. synthetic classification with calibration-aware metrics.

This is a repository of transformer-like models, including Transformer, GPT, BERT, ViT and much more to be implemented along my journey into the fascinating deep learning field.

CLIP: Connecting Text and Image (Learning Transferable Visual Models From Natural Language Supervision)

Improving Chest X-Ray Report Generation by Leveraging Warm-Starting

Code for Video Deepfake Detector from "MINTIME: Multi-Identity Size-Invariant Video Deepfake Detection", paper available on IEEE Transactions on Information Forensics and Security.

Awesome UNet with Transformer

Vision Transformers for image classification, image segmentation, and object detection.

CVPR 2024 accepted paper, An Upload-Efficient Scheme for Transferring Knowledge From a Server-Side Pre-trained Generator to Clients in Heterogeneous Federated Learning

[IEEE TMI 2024] MultiEYE: Dataset and Benchmark for OCT-Enhanced Retinal Disease Recognition from Fundus Images

Predict the Body Mass Index with one image of a human face, with state-of-the-art results.

A jupyter notebook showing how to finetune the vision transformer on a facial expression dataset (FER-2013)

[CVPR 2025 Highlight] Meta LoRA / MetaPEFT: Meta-Learning Hyperparameters for Parameter-Efficient Fine-Tuning (LoRA, Adapter, Prompt Tuning) — Automatic hyperparameter optimization via bi-level meta-learning

Sparse Autoencoders (SAE) vs CLIP fine-tuning fun.

Official PyTorch Implementation for the "Learning on Model Weights using Tree Experts" paper (CVPR 2025).

Python script to fine tune Open source Video Vision Transformer (ViVit) using HuggingFace Trainer Library

A web app using Vision Transformers for quick and accurate diagnosis of plant diseases, aiding effective treatment and plant health.

Computer Vision project focused on detecting smoke and fire in wild environments. The Google Vision Transformer was fine-tuned on a custom dataset.

Landcover classification on sentinel-2 data with Prithvi, EfficientNet-Unet and OSM / CNES Landcover labels.