Mobile-friendly Image de-noising: Hardware Conscious Optimization for Edge Application

• URL: http://arxiv.org/abs/2601.11684v1
• Date: Fri, 16 Jan 2026 09:39:01 GMT
• Title: Mobile-friendly Image de-noising: Hardware Conscious Optimization for Edge Application
• Abstract summary: This work presents a novel mobile-friendly network for image de-noising obtained with Entropy-Regularized differentiable Neural Architecture Search (NAS)<n>The designed model has 12% less parameters, with 2-fold improvement in ondevice latency and 1.5-fold improvement in the memory footprint for a 0.7% drop in PSNR, when deployed and profiled on Samsung Galaxy S24 Ultra.
• Score: 1.5658704610960574
• License: http://creativecommons.org/licenses/by/4.0/
• Abstract: Image enhancement is a critical task in computer vision and photography that is often entangled with noise. This renders the traditional Image Signal Processing (ISP) ineffective compared to the advances in deep learning. However, the success of such methods is increasingly associated with the ease of their deployment on edge devices, such as smartphones. This work presents a novel mobile-friendly network for image de-noising obtained with Entropy-Regularized differentiable Neural Architecture Search (NAS) on a hardware-aware search space for a U-Net architecture, which is first-of-its-kind. The designed model has 12% less parameters, with ~2-fold improvement in ondevice latency and 1.5-fold improvement in the memory footprint for a 0.7% drop in PSNR, when deployed and profiled on Samsung Galaxy S24 Ultra. Compared to the SOTA Swin-Transformer for Image Restoration, the proposed network had competitive accuracy with ~18-fold reduction in GMACs. Further, the network was tested successfully for Gaussian de-noising with 3 intensities on 4 benchmarks and real-world de-noising on 1 benchmark demonstrating its generalization ability.

Related papers

• Pre-training Enables Extraordinary All-optical Image Denoising [14.509895751629578]
  This paper reports a pre-training-driven approach that leads to snapshot image denoising with substantially improved quality.<n>We demonstrated effective free-space optical denoising by a diffractive network optimized by a two-step process.<n>We further demonstrated the critical role of our optical denoisers in vision-based applications, including face detection, plate recognition, and localization of UAVs in noisy conditions.
  arXiv  Detail & Related papers  (2026-05-08T14:43:54Z)
• High-resolution Photo Enhancement in Real-time: A Laplacian Pyramid Network [73.19214585791268]
  This paper introduces a pyramid network called LLF-LUT++, which integrates global and local operators through closed-form Laplacian pyramid decomposition and reconstruction.<n>Specifically, we utilize an image-adaptive 3D LUT that capitalizes on the global tonal characteristics of downsampled images.<n>LLF-LUT++ not only achieves a 2.64 dB improvement in PSNR on the HDR+ dataset, but also further reduces, with 4K resolution images processed in just 13 ms on a single GPU.
  arXiv  Detail & Related papers  (2025-10-13T16:52:32Z)
• MoiréNet: A Compact Dual-Domain Network for Image Demoiréing [14.959312694378765]
  Moir'e patterns arise from spectral aliasing between display pixel lattices and camera sensor grids.<n>We propose Moir'eNet, a convolutional neural U-Net-based framework that integrates frequency and spatial domain features for effective artifact removal.
  arXiv  Detail & Related papers  (2025-09-23T12:33:23Z)
• DnLUT: Ultra-Efficient Color Image Denoising via Channel-Aware Lookup Tables [60.95483707212802]
  DnLUT is an ultra-efficient lookup table-based framework that achieves high-quality color image denoising with minimal resource consumption.<n>Our key innovation lies in two complementary components: a Pairwise Channel Mixer (PCM) that effectively captures inter-channel correlations and spatial dependencies in parallel, and a novel L-shaped convolution design that maximizes receptive field coverage.<n>By converting these components into optimized lookup tables post-training, DnLUT achieves remarkable efficiency - requiring only 500KB storage and 0.1% energy consumption compared to its CNN contestant DnCNN, while delivering 20X faster inference.
  arXiv  Detail & Related papers  (2025-03-20T08:15:29Z)
• Practical Blind Image Denoising via Swin-Conv-UNet and Data Synthesis [148.16279746287452]
  We propose a swin-conv block to incorporate the local modeling ability of residual convolutional layer and non-local modeling ability of swin transformer block. For the training data synthesis, we design a practical noise degradation model which takes into consideration different kinds of noise. Experiments on AGWN removal and real image denoising demonstrate that the new network architecture design achieves state-of-the-art performance.
  arXiv  Detail & Related papers  (2022-03-24T18:11:31Z)
• Selective Residual M-Net for Real Image Denoising [6.909688694501238]
  We propose a blind real image denoising network (SRMNet) to advance the performance of denoising algorithms. Specifically, we use a selective kernel with residual block on the hierarchical structure called M-Net to enrich the multi-scale semantic information. OurNet has competitive performance results on two synthetic and two real-world noisy datasets in terms of quantitative metrics and visual quality.
  arXiv  Detail & Related papers  (2022-03-03T11:10:30Z)
• A New Backbone for Hyperspectral Image Reconstruction [90.48427561874402]
  3D hyperspectral image (HSI) reconstruction refers to inverse process of snapshot compressive imaging. Proposal is for a Spatial/Spectral Invariant Residual U-Net, namely SSI-ResU-Net. We show that SSI-ResU-Net achieves competing performance with over 77.3% reduction in terms of floating-point operations.
  arXiv  Detail & Related papers  (2021-08-17T16:20:51Z)
• Feature-Align Network and Knowledge Distillation for Efficient Denoising [4.997028216419175]
  Deep learning-based RAW image denoising is a quintessential problem in image restoration. We propose a novel network for efficient RAW denoising on mobile devices. We achieve a PSNR of 48.28dB, with 263 times fewer MACs, and 17.6 times fewer parameters than the state-of-the-art network.
  arXiv  Detail & Related papers  (2021-03-02T07:09:32Z)
• Neighbor2Neighbor: Self-Supervised Denoising from Single Noisy Images [98.82804259905478]
  We present Neighbor2Neighbor to train an effective image denoising model with only noisy images. In detail, input and target used to train a network are images sub-sampled from the same noisy image. A denoising network is trained on sub-sampled training pairs generated in the first stage, with a proposed regularizer as additional loss for better performance.
  arXiv  Detail & Related papers  (2021-01-08T02:03:25Z)
• Practical Deep Raw Image Denoising on Mobile Devices [30.3578422624862]
  We propose a light-weight, efficient neural network-based raw image denoiser that runs smoothly on mainstream mobile devices. Our proposed mobile-friendly denoising model runs at 70 milliseconds per megapixel on Qualcomm Snapdragon 855 chipset.
  arXiv  Detail & Related papers  (2020-10-14T10:30:32Z)
• Attention Based Real Image Restoration [48.933507352496726]
  Deep convolutional neural networks perform better on images containing synthetic degradations. This paper proposes a novel single-stage blind real image restoration network (R$2$Net)
  arXiv  Detail & Related papers  (2020-04-26T04:21:49Z)
• Pre-defined Sparsity for Low-Complexity Convolutional Neural Networks [9.409651543514615]
  This work introduces convolutional layers with pre-defined sparse 2D kernels that have support sets that repeat periodically within and across filters. Due to the efficient storage of our periodic sparse kernels, the parameter savings can translate into considerable improvements in energy efficiency.
  arXiv  Detail & Related papers  (2020-01-29T07:10:56Z)

This list is automatically generated from the titles and abstracts of the papers in this site.

This site does not guarantee the quality of this site (including all information) and is not responsible for any consequences.