Abstract: Digital camera sensors may struggle to capture images in low-light environments, resulting in lower brightness and contrast levels, color degradation, undesirable characteristics, and noise.

Abstract: Graph Convolution Networks (GCNs) have achieved remarkable success in representation of structured graph data. As we know that traditional GCNs are generally defined on the fixed first-order ...

The rapid progress of large language models (LLMs) has catalyzed the emergence of multimodal large language models (MLLMs) that unify visual understanding and image generation within a single ...

Integrates dynamic codebook frequency statistics into a transformer attention module. Fuses semantic image features with latent representations of quantization ...