🏢 Huazhong University of Science and Technology

This survey explores the evolution of physics cognition in video generation, addressing the gap between visual realism and physical accuracy.

Free4D: Tuning-free 4D scene generation with spatial-temporal consistency.

LLMs modestly affect Wikipedia, subtly altering content and potentially skewing NLP benchmarks.

CROWDSELECT boosts instruction tuning by cleverly selecting synthetic data using multi-LLM wisdom, enhancing model performance across diverse tasks.

LightningDiT resolves the optimization dilemma in latent diffusion models by aligning latent space with pre-trained vision models, achieving state-of-the-art ImageNet 256x256 generation with over 21x …