Search papers, labs, and topics across Lattice.
This paper introduces DiMOO-SR, a rarity-aware multimodal discrete diffusion framework designed to enhance photo-realistic image super-resolution (SR) by addressing the challenges of long-tailed token distributions and spatial inconsistencies in parallel decoding. By employing Inverse Frequency Sampling (IFS) during training to prioritize rare yet critical textures and Spatial Consistency Ranking (SCR) during inference to ensure structural coherence, the method achieves competitive perceptual quality with significantly fewer decoding steps. The results demonstrate the effectiveness of discrete diffusion as a viable approach for generative image SR, offering a promising alternative to traditional continuous models.
Rarity-aware sampling and spatial consistency techniques enable discrete diffusion to achieve high-quality image super-resolution with fewer decoding steps than traditional methods.
Continuous diffusion models have become the dominant paradigm for photo-realistic image Super-Resolution (SR), but they typically formulate reconstruction as continuous signal-level denoising and incorporate semantic priors through external conditioning modules. This makes it less direct to exploit the unified token-based scaling paradigm of modern multimodal models. Autoregressive models provide a more native semantic representation by modeling images as discrete visual tokens, yet their causal decoding is inefficient for high-resolution reconstruction. Discrete diffusion offers a promising middle ground by enabling non-causal, parallel prediction over visual tokens. However, directly adapting discrete diffusion to SR remains non-trivial due to two task-specific challenges: (1) the long-tailed distribution of visual tokens, which under-represents rare but perceptually critical textures; and (2) spatially inconsistent parallel decoding, which may introduce isolated artifacts. To address these issues, we propose DiMOO-SR, a rarity-aware multimodal discrete diffusion framework for photo-realistic image SR. During training, Inverse Frequency Sampling (IFS) prioritizes under-represented but information-rich tokens. During inference, Spatial Consistency Ranking (SCR) refines token confidence using local neighborhood agreement to improve structural coherence. Extensive experiments on widely used real-world SR benchmarks demonstrate that DiMOO-SR achieves competitive perceptual quality with only a few parallel decoding steps, highlighting the potential of discrete diffusion for generative image super-resolution. The code will be released upon publication.
