dinov2

A microframework on top of PyTorch with first-class citizen APIs for foundation model adaptation

[ECCV 2024] Improving 2D Feature Representations by 3D-Aware Fine-Tuning

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.