A Perspective on Deep Vision Performance with Standard Image and Video Codecs
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_ | 1866910636302139392 |
|---|---|
| author | Reich, Christoph Hahn, Oliver Cremers, Daniel Roth, Stefan Debnath, Biplob |
| author_facet | Reich, Christoph Hahn, Oliver Cremers, Daniel Roth, Stefan Debnath, Biplob |
| contents | Resource-constrained hardware, such as edge devices or cell phones, often rely on cloud servers to provide the required computational resources for inference in deep vision models. However, transferring image and video data from an edge or mobile device to a cloud server requires coding to deal with network constraints. The use of standardized codecs, such as JPEG or H.264, is prevalent and required to ensure interoperability. This paper aims to examine the implications of employing standardized codecs within deep vision pipelines. We find that using JPEG and H.264 coding significantly deteriorates the accuracy across a broad range of vision tasks and models. For instance, strong compression rates reduce semantic segmentation accuracy by more than 80% in mIoU. In contrast to previous findings, our analysis extends beyond image and action classification to localization and dense prediction tasks, thus providing a more comprehensive perspective. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2404_12330 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | A Perspective on Deep Vision Performance with Standard Image and Video Codecs Reich, Christoph Hahn, Oliver Cremers, Daniel Roth, Stefan Debnath, Biplob Computer Vision and Pattern Recognition Multimedia Resource-constrained hardware, such as edge devices or cell phones, often rely on cloud servers to provide the required computational resources for inference in deep vision models. However, transferring image and video data from an edge or mobile device to a cloud server requires coding to deal with network constraints. The use of standardized codecs, such as JPEG or H.264, is prevalent and required to ensure interoperability. This paper aims to examine the implications of employing standardized codecs within deep vision pipelines. We find that using JPEG and H.264 coding significantly deteriorates the accuracy across a broad range of vision tasks and models. For instance, strong compression rates reduce semantic segmentation accuracy by more than 80% in mIoU. In contrast to previous findings, our analysis extends beyond image and action classification to localization and dense prediction tasks, thus providing a more comprehensive perspective. |
| title | A Perspective on Deep Vision Performance with Standard Image and Video Codecs |
| topic | Computer Vision and Pattern Recognition Multimedia |
| url | https://arxiv.org/abs/2404.12330 |