Serpent: Scalable and Efficient Image Restoration via Multi-scale Structured State Space Models
Fuente:
arXiv
Salvato in:
| Autori principali: | , , |
|---|---|
| Natura: | Preprint |
| Pubblicazione: |
2024
|
| Soggetti: | |
| Accesso online: | |
| Tags: |
Aggiungi Tag
Nessun Tag, puoi essere il primo ad aggiungerne!!
|
| _version_ | 1866909463495049216 |
|---|---|
| author | Sepehri, Mohammad Shahab Fabian, Zalan Soltanolkotabi, Mahdi |
| author_facet | Sepehri, Mohammad Shahab Fabian, Zalan Soltanolkotabi, Mahdi |
| contents | The landscape of computational building blocks of efficient image restoration architectures is dominated by a combination of convolutional processing and various attention mechanisms. However, convolutional filters, while efficient, are inherently local and therefore struggle with modeling long-range dependencies in images. In contrast, attention excels at capturing global interactions between arbitrary image regions, but suffers from a quadratic cost in image dimension. In this work, we propose Serpent, an efficient architecture for high-resolution image restoration that combines recent advances in state space models (SSMs) with multi-scale signal processing in its core computational block. SSMs, originally introduced for sequence modeling, can maintain a global receptive field with a favorable linear scaling in input size. We propose a novel hierarchical architecture inspired by traditional signal processing principles, that converts the input image into a collection of sequences and processes them in a multi-scale fashion. Our experimental results demonstrate that Serpent can achieve reconstruction quality on par with state-of-the-art techniques, while requiring orders of magnitude less compute (up to $150$ fold reduction in FLOPS) and a factor of up to $5\times$ less GPU memory while maintaining a compact model size. The efficiency gains achieved by Serpent are especially notable at high image resolutions. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2403_17902 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | Serpent: Scalable and Efficient Image Restoration via Multi-scale Structured State Space Models Sepehri, Mohammad Shahab Fabian, Zalan Soltanolkotabi, Mahdi Image and Video Processing Computer Vision and Pattern Recognition Machine Learning I.4.4; I.4.5 The landscape of computational building blocks of efficient image restoration architectures is dominated by a combination of convolutional processing and various attention mechanisms. However, convolutional filters, while efficient, are inherently local and therefore struggle with modeling long-range dependencies in images. In contrast, attention excels at capturing global interactions between arbitrary image regions, but suffers from a quadratic cost in image dimension. In this work, we propose Serpent, an efficient architecture for high-resolution image restoration that combines recent advances in state space models (SSMs) with multi-scale signal processing in its core computational block. SSMs, originally introduced for sequence modeling, can maintain a global receptive field with a favorable linear scaling in input size. We propose a novel hierarchical architecture inspired by traditional signal processing principles, that converts the input image into a collection of sequences and processes them in a multi-scale fashion. Our experimental results demonstrate that Serpent can achieve reconstruction quality on par with state-of-the-art techniques, while requiring orders of magnitude less compute (up to $150$ fold reduction in FLOPS) and a factor of up to $5\times$ less GPU memory while maintaining a compact model size. The efficiency gains achieved by Serpent are especially notable at high image resolutions. |
| title | Serpent: Scalable and Efficient Image Restoration via Multi-scale Structured State Space Models |
| topic | Image and Video Processing Computer Vision and Pattern Recognition Machine Learning I.4.4; I.4.5 |
| url | https://arxiv.org/abs/2403.17902 |