Responsive DNN Adaptation for Video Analytics against Environment Shift via Hierarchical Mobile-Cloud Collaborations

2025-04-30

AI Summary

To address the sluggish adaptation of expert DNN models and degraded cloud-centric performance under dynamic environmental changes in mobile video analytics, this paper proposes MOCHA, a mobile-cloud hierarchical collaborative framework. Methodologically, MOCHA introduces a novel co-design of lightweight on-device model reuse and cloud-side semantic indexing retrieval; establishes a foundation-model-driven taxonomy for expert models; and devises an active prefetching caching strategy. It integrates on-device fast fine-tuning, cloud-side semantic analysis, structured model indexing, and demand-driven prefetching. Evaluated across three DNN tasks, MOCHA improves adaptation-phase accuracy by up to 6.8%, reduces response latency to 1/35.5 of the baseline, and cuts retraining time by 3.0×—significantly enhancing real-time adaptability in dynamic environments.

Technology Category

• Machine Learning: Mixture of Experts (MoE)
• Computer Vision: Multi-modal Vision
• Application Domains: Mobility, Driving & Flight

Application Category

• Systems and Infrastructure for Web, Mobile and WoT: Applied ML and AI for Web-based mobile applications
• User Modeling, Personalization and Recommendation: On-Device user modeling, personalization, and recommendation
• Search and Retrieval-Augmented AI: Web learning to rank, online learning, and counterfactual learning for ranking

Abstract

Mobile video analysis systems often encounter various deploying environments, where environment shifts present greater demands for responsiveness in adaptations of deployed"expert DNN models". Existing model adaptation frameworks primarily operate in a cloud-centric way, exhibiting degraded performance during adaptation and delayed reactions to environment shifts. Instead, this paper proposes MOCHA, a novel framework optimizing the responsiveness of continuous model adaptation through hierarchical collaborations between mobile and cloud resources. Specifically, MOCHA (1) reduces adaptation response delays by performing on-device model reuse and fast fine-tuning before requesting cloud model retrieval and end-to-end retraining; (2) accelerates history expert model retrieval by organizing them into a structured taxonomy utilizing domain semantics analyzed by a cloud foundation model as indices; (3) enables efficient local model reuse by maintaining onboard expert model caches for frequent scenes, which proactively prefetch model weights from the cloud model database. Extensive evaluations with real-world videos on three DNN tasks show MOCHA improves the model accuracy during adaptation by up to 6.8% while saving the response delay and retraining time by up to 35.5x and 3.0x respectively.

Problem

• Optimizes DNN adaptation responsiveness via mobile-cloud collaboration
• Reduces delays with on-device reuse and fast fine-tuning
• Accelerates model retrieval using domain semantics taxonomy

Innovation

• Hierarchical mobile-cloud collaboration for DNN adaptation
• On-device model reuse and fast fine-tuning
• Structured taxonomy for expert model retrieval