Superkernel Neural Architecture Search for Image Denoising

Możejko, Marcin; Latkowski, Tomasz; Treszczotko, Lukasz; Michał, Szafraniuk,; Trojanowski, Krzysztof

doi:10.1109/cvprw50498.2020.00250

Cited by 16 publications

(6 citation statements)

References 18 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…5. Efficient NAS for Image Denoising:** Developments in NAS have also been applied to image denoising, with techniques like superkernel implementations enabling fast training of models for dense predictions, showcasing NAS's versatility in different application domains [18]. 6.…”

Section: Development and Elaborationmentioning

confidence: 99%

Evolution and Efficiency in Neural Architecture Search: Bridging the Gap between Expert Design and Automated Optimization

2024

J Math Techniques Comput Math

View full text Add to dashboard Cite

The paper provides a comprehensive overview of Neural Architecture Search (NAS), emphasizing its evolution from manual design to automated, computationally driven approaches. It covers the inception and growth of NAS, highlighting its application across various domains, including medical imaging and natural language processing. The document details the shift from expert-driven design to algorithm-driven processes, exploring initial methodologies like reinforcement learning and evolutionary algorithms. It also discusses the challenges of computational demands and the emergence of efficient NAS methodologies, such as Differentiable Architecture Search and hardware-aware NAS. The paper further elaborates on NAS's application in computer vision, NLP, and beyond, demonstrating its versatility and potential for optimizing neural network architectures across different tasks. Future directions and challenges, including computational efficiency and the integration with emerging AI domains, are addressed, showcasing NAS's dynamic nature and its continued evolution towards more sophisticated and efficient architecture search methods.

show abstract

Section: Development and Elaborationmentioning

confidence: 99%

Evolution and Efficiency in Neural Architecture Search: Bridging the Gap between Expert Design and Automated Optimization

2024

J Math Techniques Comput Math

View full text Add to dashboard Cite

show abstract

“…Xu et al [21] proposed a frequency-based decomposition model to enhance the effect of low-light image enhancement by restoring image objects in low-frequency layers and enhancing details in high-frequency layers. In recent years, neural structure search techniques have gradually become a popular method for solving low-level vision tasks [22][23][24] due to their ability to potentially explore structures superior to human prior knowledge. Aiming to break the limitation of existing image restoration methods that target one degradation factor only, Li et al [22] proposed to use neural architecture search (NAS) to discover a single network that can handle multiple degradation factors simultaneously.…”

Section: Related Workmentioning

confidence: 99%

Toward Robust and Efficient Low-Light Image Enhancement: Progressive Attentive Retinex Architecture Search

Shang

Zhang

et al. 2023

Tsinghua Sci. Technol.

View full text Add to dashboard Cite

In recent years, learning-based low-light image enhancement methods have shown excellent performance, but the heuristic design adopted by most methods requires high engineering skills for developers, causing expensive inference costs that are unfriendly to the hardware platform. To handle this issue, we propose to automatically discover an efficient architecture, called progressive attentive Retinex network (PAR-Net). We define a new attentive Retinex framework by introducing the attention mechanism to strengthen structural representation. A multi-level search space containing micro-level on the operation and macro-level on the cell is established to realize meticulous construction. To endow the searched architecture with the hardware-aware property, we develop a latency-constrained progressive search strategy that successfully improves the model capability by explicitly expressing the intrinsic relationship between different models defined in the attentive Retinex framework. Extensive quantitative and qualitative experimental results fully justify the superiority of our proposed approach against other state-of-the-art methods. A series of analytical evaluations is performed to illustrate the validity of our proposed algorithm.

show abstract

“…The above search algorithms have achieved highly competitive performance in various high-level vision tasks, including image classification [36], object detection [30], and semantic segmentation. Very recently, NAS algorithms [21,41,14] have also been applied to low-level vision problems, including image denoising, restoration, and deraining, etc. Unfortunately, existing NAS strategies are fully data-driven, thus require a large number of well-prepared paired training data, which is generally impractical for the LIE.…”

Section: Neural Architecture Search (Nas)mentioning

confidence: 99%

Retinex-inspired Unrolling with Cooperative Prior Architecture Search for Low-light Image Enhancement

Liu¹,

Ma²,

Zhang³

et al. 2020

Preprint

View full text Add to dashboard Cite

Low-light image enhancement plays very important roles in low-level vision field. Recent works have built a large variety of deep learning models to address this task. However, these approaches mostly rely on significant architecture engineering and suffer from high computational burden. In this paper, we propose a new method, named Retinexinspired Unrolling with Architecture Search (RUAS), to construct lightweight yet effective enhancement network for low-light images in real-world scenario. Specifically, building upon Retinex rule, RUAS first establishes models to characterize the intrinsic underexposed structure of lowlight images and unroll their optimization processes to construct our holistic propagation structure. Then by designing a cooperative reference-free learning strategy to discover low-light prior architectures from a compact search space, RUAS is able to obtain a top-performing image enhancement network, which is with fast speed and requires few computational resources. Extensive experiments verify the superiority of our RUAS framework against recently proposed state-of-the-art methods.

show abstract

Superkernel Neural Architecture Search for Image Denoising

Cited by 16 publications

References 18 publications

Evolution and Efficiency in Neural Architecture Search: Bridging the Gap between Expert Design and Automated Optimization

Evolution and Efficiency in Neural Architecture Search: Bridging the Gap between Expert Design and Automated Optimization

Toward Robust and Efficient Low-Light Image Enhancement: Progressive Attentive Retinex Architecture Search

Retinex-inspired Unrolling with Cooperative Prior Architecture Search for Low-light Image Enhancement

Contact Info

Product

Resources

About