comprehensive evaluation
A thorough assessment methodology for AI models that considers multiple performance metrics, including accuracy, speed, data efficiency, and robustness, ensuring a holistic understanding of a model’s capabilities and limitations.
- BackdoorDM: A Comprehensive Benchmark for Backdoor Learning on Diffusion Model
- Comprehensive Assessment and Analysis for NSFW Content Erasure in Text-to-Image Diffusion models
- E2E-VGuard: Adversarial Prevention for Production LLM-based End-To-End Speech Synthesis
- Guard Me If You Know Me: Protecting Specific Face-Identity from Deepfakes
- MME: A Comprehensive Evaluation Benchmark for Multimodal Large Language Models
- MUVR: A Multi-Modal Untrimmed Video Retrieval Benchmark with Multi-Level Visual Correspondence
- Mitigating Hallucination Through Theory-Consistent Symmetric Multimodal Preference Optimization
- ROSE: Remove Objects with Side Effects in Videos
- SafeVid: Toward Safety Aligned Video Large Multimodal Models
- The Fluorescent Veil: A Stealthy and Effective Physical Adversarial Patch Against Traffic Sign Recognition
- The Future Unmarked: Watermark Removal in AI-Generated Images via Next-Frame Prediction
- Towards Reliable Identification of Diffusion-based Image Manipulations
- Trans-EnV: A Framework for Evaluating the Linguistic Robustness of LLMs Against English Varieties