Saved in:
| Main Authors: | , , , |
|---|---|
| Format: | Preprint |
| Published: |
2025
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2509.16853 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866909981954015232 |
|---|---|
| author | Wang, Jinhao Ling, Nam Wang, Wei Jiang, Wei |
| author_facet | Wang, Jinhao Ling, Nam Wang, Wei Jiang, Wei |
| contents | Prior studies in embodied AI consistently show that robust perception is critical for human-robot interaction, yet deploying high-fidelity visual models on resource-constrained agents remains challenging due to limited on-device computation power and transmission latency. Exploiting the redundancy in latent representations could improve system efficiency, yet existing approaches often rely on costly dataset-specific ablation tests or heavy entropy models unsuitable for real-time edge-robot collaboration.
We propose a generalizable, dataset-agnostic method to identify and selectively transmit structure-critical channels in pretrained encoders. Instead of brute-force empirical evaluations, our approach leverages intrinsic parameter statistics-weight variances and biases-to estimate channel importance. This analysis reveals a consistent organizational structure, termed the Invariant Salient Channel Space (ISCS), where Salient-Core channels capture dominant structures while Salient-Auxiliary channels encode fine visual details. Building on ISCS, we introduce a deterministic static pruning strategy that enables lightweight split-computing.
Experiments across different datasets demonstrate that our method achieves a deterministic, ultra-low latency pipeline by bypassing heavy entropy modeling. Our method reduces end-to-end latency, providing a critical speed-accuracy trade-off for resource-constrained human-aware embodied systems. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2509_16853 |
| institution | arXiv |
| publishDate | 2025 |
| record_format | arxiv |
| spellingShingle | ISCS: Parameter-Guided Feature Pruning for Resource-Constrained Embodied Perception Wang, Jinhao Ling, Nam Wang, Wei Jiang, Wei Computer Vision and Pattern Recognition Prior studies in embodied AI consistently show that robust perception is critical for human-robot interaction, yet deploying high-fidelity visual models on resource-constrained agents remains challenging due to limited on-device computation power and transmission latency. Exploiting the redundancy in latent representations could improve system efficiency, yet existing approaches often rely on costly dataset-specific ablation tests or heavy entropy models unsuitable for real-time edge-robot collaboration. We propose a generalizable, dataset-agnostic method to identify and selectively transmit structure-critical channels in pretrained encoders. Instead of brute-force empirical evaluations, our approach leverages intrinsic parameter statistics-weight variances and biases-to estimate channel importance. This analysis reveals a consistent organizational structure, termed the Invariant Salient Channel Space (ISCS), where Salient-Core channels capture dominant structures while Salient-Auxiliary channels encode fine visual details. Building on ISCS, we introduce a deterministic static pruning strategy that enables lightweight split-computing. Experiments across different datasets demonstrate that our method achieves a deterministic, ultra-low latency pipeline by bypassing heavy entropy modeling. Our method reduces end-to-end latency, providing a critical speed-accuracy trade-off for resource-constrained human-aware embodied systems. |
| title | ISCS: Parameter-Guided Feature Pruning for Resource-Constrained Embodied Perception |
| topic | Computer Vision and Pattern Recognition |
| url | https://arxiv.org/abs/2509.16853 |