# Covariance-Guided Resource Adaptive Learning for Efficient Edge Inference

arXiv: https://arxiv.org/abs/2603.14577 · HTML: https://mlsanity.com/paper/2603.14577/

## Key facts

- No citing papers in the map yet (data through 14 Oct 2026).
- Builds on 9 of its 22 references that are in the map.
- First posted on arXiv 15 Mar 2026.

## Builds on (most cited first, 9 of 9)

- [Deep Residual Learning for Image Recognition](https://mlsanity.com/paper/1512.03385/) (2015)
- [Microsoft COCO: Common Objects in Context](https://mlsanity.com/paper/1405.0312/) (2014)
- [Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networks](https://mlsanity.com/paper/1506.01497/) (2015)
- [Searching for MobileNetV3](https://mlsanity.com/paper/1905.02244/) (2019)
- [DVFO: Learning-Based DVFS for Energy-Efficient Edge-Cloud Collaborative Inference](https://mlsanity.com/paper/2306.01811/) (2023)
- [Miriam: Exploiting Elastic Kernels for Real-time Multi-DNN Inference on Edge GPU](https://mlsanity.com/paper/2307.04339/) (2023)
- [Shared Memory-contention-aware Concurrent DNN Execution for Diversely Heterogeneous System-on-Chips](https://mlsanity.com/paper/2308.05869/) (2023)
- [PolyThrottle: Energy-efficient Neural Network Inference on Edge Devices](https://mlsanity.com/paper/2310.19991/) (2023)
- [AI Multi-Tenancy on Edge: Concurrent Deep Learning Model Executions and Dynamic Model Placements on Edge Devices](https://mlsanity.com/paper/2107.12486/) (2021)

## Abstract (arXiv)

For deep learning inference on edge devices, hardware configurations achieving the same throughput can differ by 2× in power consumption, yet operators often struggle to find the efficient ones without exhaustive profiling. Existing approaches often rely on inefficient static presets or require expensive offline profiling that must be repeated for each new model or device. To address this problem, we present CORAL, an online optimization method that discovers near-optimal configurations without offline profiling. CORAL leverages distance covariance to statistically capture the non-linear dependencies between hardware settings, e.g., DVFS and concurrency levels, and performance metrics. Unlike prior work, we explicitly formulate the challenge as a throughput-power co-optimization problem to satisfy power budgets and throughput targets simultaneously. We evaluate CORAL on two NVIDIA Jetson devices across three object detection models ranging from lightweight to heavyweight. In single-target scenarios, CORAL achieves 96% unicodex2013 100% of the optimal performance found by exhaustive search. In strict dual-constraint scenarios where baselines fail or exceed power budgets, CORAL consistently finds proper configurations online with minimal exploration.