Search papers, labs, and topics across Lattice.
2
0
4
4
Ditch the camera metadata: this diffusion model learns to generate realistic sRGB noise directly from noisy image prompts, boosting denoising performance in real-world scenarios.
LLM-augmented training with similarity-aware masking lets weakly-supervised video captioning models generate more accurate event descriptions and temporal boundaries, even with sparse training data.
