Saved in:
| Main Authors: | , , , , , |
|---|---|
| Format: | Preprint |
| Published: |
2024
|
| Subjects: | |
| Online Access: | https://arxiv.org/abs/2409.11403 |
| Tags: |
Add Tag
No Tags, Be the first to tag this record!
|
| _version_ | 1866918171594719232 |
|---|---|
| author | Sengupta, Kathakoli Shangguan, Zhongkai Bharadwaj, Sandesh Arora, Sanjay Ohn-Bar, Eshed Mancuso, Renato |
| author_facet | Sengupta, Kathakoli Shangguan, Zhongkai Bharadwaj, Sandesh Arora, Sanjay Ohn-Bar, Eshed Mancuso, Renato |
| contents | Embodied vision-based real-world systems, such as mobile robots, require a careful balance between energy consumption, compute latency, and safety constraints to optimize operation across dynamic tasks and contexts. As local computation tends to be restricted, offloading the computation, ie, to a remote server, can save local resources while providing access to high-quality predictions from powerful and large models. However, the resulting communication and latency overhead has led to limited usability of cloud models in dynamic, safety-critical, real-time settings. To effectively address this trade-off, we introduce UniLCD, a novel hybrid inference framework for enabling flexible local-cloud collaboration. By efficiently optimizing a flexible routing module via reinforcement learning and a suitable multi-task objective, UniLCD is specifically designed to support the multiple constraints of safety-critical end-to-end mobile systems. We validate the proposed approach using a challenging, crowded navigation task requiring frequent and timely switching between local and cloud operations. UniLCD demonstrates improved overall performance and efficiency, by over 35% compared to state-of-the-art baselines based on various split computing and early exit strategies. |
| format | Preprint |
| id |
arxiv_https___arxiv_org_abs_2409_11403 |
| institution | arXiv |
| publishDate | 2024 |
| record_format | arxiv |
| spellingShingle | UniLCD: Unified Local-Cloud Decision-Making via Reinforcement Learning Sengupta, Kathakoli Shangguan, Zhongkai Bharadwaj, Sandesh Arora, Sanjay Ohn-Bar, Eshed Mancuso, Renato Robotics Embodied vision-based real-world systems, such as mobile robots, require a careful balance between energy consumption, compute latency, and safety constraints to optimize operation across dynamic tasks and contexts. As local computation tends to be restricted, offloading the computation, ie, to a remote server, can save local resources while providing access to high-quality predictions from powerful and large models. However, the resulting communication and latency overhead has led to limited usability of cloud models in dynamic, safety-critical, real-time settings. To effectively address this trade-off, we introduce UniLCD, a novel hybrid inference framework for enabling flexible local-cloud collaboration. By efficiently optimizing a flexible routing module via reinforcement learning and a suitable multi-task objective, UniLCD is specifically designed to support the multiple constraints of safety-critical end-to-end mobile systems. We validate the proposed approach using a challenging, crowded navigation task requiring frequent and timely switching between local and cloud operations. UniLCD demonstrates improved overall performance and efficiency, by over 35% compared to state-of-the-art baselines based on various split computing and early exit strategies. |
| title | UniLCD: Unified Local-Cloud Decision-Making via Reinforcement Learning |
| topic | Robotics |
| url | https://arxiv.org/abs/2409.11403 |