Breaking the accuracy-resource dilemma: a lightweight adaptive video inference enhancement
Fuente:
arXiv
Saved in:
| Main Authors: | , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2026
|
| Subjects: | |
| Online Access: | |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866916019220512768 |
|---|---|
| author | Ma, Wei Chen, Shaowu Ye, Junjie Zhang, Peichang Huang, Lei |
| author_facet | Ma, Wei Chen, Shaowu Ye, Junjie Zhang, Peichang Huang, Lei |
| contents | Existing video inference (VI) enhancement methods typically aim to improve performance by scaling up model sizes and employing sophisticated network architectures. While these approaches demonstrated state-of-the-art performance, they often overlooked the trade-off of resource efficiency and inference effectiveness, leading to inefficient resource utilization and suboptimal inference performance. To address this problem, a fuzzy controller (FC-r) is developed based on key system parameters and inference-related metrics. Guided by the FC-r, a VI enhancement framework is proposed, where the spatiotemporal correlation of targets across adjacent video frames is leveraged. Given the real-time resource conditions of the target device, the framework can dynamically switch between models of varying scales during VI. Experimental results demonstrate that the proposed method effectively achieves a balance between resource utilization and inference performance. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2601_14568 |
| institution | arXiv |
| publishDate | 2026 |
| record_format | arxiv |
| spellingShingle | Breaking the accuracy-resource dilemma: a lightweight adaptive video inference enhancement Ma, Wei Chen, Shaowu Ye, Junjie Zhang, Peichang Huang, Lei Computer Vision and Pattern Recognition Artificial Intelligence Existing video inference (VI) enhancement methods typically aim to improve performance by scaling up model sizes and employing sophisticated network architectures. While these approaches demonstrated state-of-the-art performance, they often overlooked the trade-off of resource efficiency and inference effectiveness, leading to inefficient resource utilization and suboptimal inference performance. To address this problem, a fuzzy controller (FC-r) is developed based on key system parameters and inference-related metrics. Guided by the FC-r, a VI enhancement framework is proposed, where the spatiotemporal correlation of targets across adjacent video frames is leveraged. Given the real-time resource conditions of the target device, the framework can dynamically switch between models of varying scales during VI. Experimental results demonstrate that the proposed method effectively achieves a balance between resource utilization and inference performance. |
| title | Breaking the accuracy-resource dilemma: a lightweight adaptive video inference enhancement |
| topic | Computer Vision and Pattern Recognition Artificial Intelligence |
| url | https://arxiv.org/abs/2601.14568 |