# Steered Mixture of Experts Regression for Image Denoising with Multi-Model-Inference

arXiv: https://arxiv.org/abs/2303.17409 · HTML: https://mlsanity.com/paper/2303.17409/

## Key facts

- Cited by 4 papers in the map (data through 14 Oct 2026).
- Placed in Vision mixture-of-experts, one of 12 directions in Mixture-of-experts.
- Builds on 4 of its 31 references that are in the map.
- First posted on arXiv 30 Mar 2023.

## Placed in

- Mixture-of-experts / [Vision mixture-of-experts](https://mlsanity.com/moe/vision-mixture-of-experts/)

## Builds on (most cited first, 4 of 4)

- [Deep Learning on Image Denoising: An overview](https://mlsanity.com/paper/1912.13171/) (2019)
- [NTIRE 2020 Challenge on Real Image Denoising: Dataset, Methods and Results](https://mlsanity.com/paper/2005.04117/) (2020)
- [Edge-Aware Autoencoder Design for Real-Time Mixture-of-Experts Image Compression](https://mlsanity.com/paper/2207.12348/) (2022)
- [New Risk Bounds for 2D Total Variation Denoising](https://mlsanity.com/paper/1902.01215/) (2019)

## Built on by (most cited first, 4 of 4)

- [Steered Mixture-of-Experts Autoencoder Design for Real-Time Image Modelling and Denoising](https://mlsanity.com/paper/2305.03485/) (2023)
- [Adaptive Segmentation-Based Initialization for Steered Mixture of Experts Image Regression](https://mlsanity.com/paper/2409.10101/) (2024)
- [Denoising OCT Images Using Steered Mixture of Experts with Multi-Model Inference](https://mlsanity.com/paper/2402.12735/) (2024)
- [Rasterized Steered Mixture of Experts for Efficient 2D Image Regression](https://mlsanity.com/paper/2510.05814/) (2025)

## Abstract (arXiv)

In this paper we introduce a novel block-based regression strategy for image denoising based on edge-aware Steered-Mixture-of-Experts (SMoE) models. SMoEs provide very sparse image representations, able to model sharp edges as well as smooth transitions in images efficiently with few parameters. A multi-model inference strategy is developed that improves significantly the denoising capacity of single SMoE models. We show that the important edge reconstruction properties of SMoEs are well preserved, even when many models are fused under severe noise. We investigate model-inference from local neighborhood blocks as well as from distant blocks using block-matching as in BM3D. Our initial results indicate that SMoE multi-model regression can provide promising results compared to state-of-the-art BM3D with excellent edge quality.