ROMA: Run-Time Object Detection To Maximize Real-Time Accuracy
Not on the map. Why?
Abstract (arXiv)
This paper analyzes the effects of dynamically varying video contents and detection latency on the real-time detection accuracy of a detector and proposes a new run-time accuracy variation model, ROMA, based on the findings from the analysis. ROMA is designed to select an optimal detector out of a set of detectors in real time without label information to maximize real-time object detection accuracy. ROMA utilizing four YOLOv4 detectors on an NVIDIA Jetson Nano shows real-time accuracy improvements by 4 to 37% for a scenario of dynamically varying video contents and detection latency consisting of MOT17Det and MOT20Det datasets, compared to individual YOLOv4 detectors and two state-of-the-art runtime techniques.
Key facts. Cited by 2 papers in the map (data through 7 Oct 2026). Builds on 23 of its 35 references that are in the map. First posted on arXiv 28 Oct 2022.
Builds on
23 of 35 references.
References · 23
- 2015Faster R-CNN: Towards Real-Time Object Detection with Region Proposal Networkscited by 9,617
- 2019EfficientNet: Rethinking Model Scaling for Convolutional Neural Networkscited by 6,415
- 2014Unsupervised Domain Adaptation by Backpropagationcited by 1,906
- 2020YOLOv4: Optimal Speed and Accuracy of Object Detectioncited by 1,624
- 2016Neural Architecture Search with Reinforcement Learningcited by 1,489
- 2019EfficientDet: Scalable and Efficient Object Detectioncited by 1,018
- 2018MnasNet: Platform-Aware Neural Architecture Search for Mobilecited by 937
- 2019Once-for-All: Train One Network and Specialize it for Efficient Deploymentcited by 778
- 2018FBNet: Hardware-Aware Efficient ConvNet Design via Differentiable Neural Architecture Searchcited by 480
- 2018AMC: AutoML for Model Compression and Acceleration on Mobile Devicescited by 375
- 2016Domain Separation Networkscited by 317
- 2019Towards Real-Time Multi-Object Trackingcited by 311
11 more
- 2018Slimmable Neural Networkscited by 301
- 2017NISP: Pruning Networks using Neuron Importance Score Propagationcited by 228
- 2016Speed/accuracy trade-offs for modern convolutional object detectorscited by 198
- 2020Can Autonomous Vehicles Identify, Recover From, and Adapt to Distribution Shifts?cited by 139
- 2015MOTChallenge 2015: Towards a Benchmark for Multi-Target Trackingcited by 133
- 2018NestDNN: Resource-Aware Multi-Tenant On-Device Deep Learning for Continuous Mobile Visioncited by 68
- 2021Dynamic-OFA: Runtime DNN Architecture Switching for Performance Scaling on Heterogeneous Embedded Platformscited by 18
- 2020ApproxDet: Content and Contention-Aware Approximate Object Detection for Mobilescited by 15
- 2019Optimizing Deep Learning Inference on Embedded Systems Through Adaptive Model Selectioncited by 6
- 2020Performance-Oriented Neural Architecture Searchcited by 3
- 2021TOD: Transprecise Object Detection to Maximise Real-Time Accuracy on the Edgecited by 3
Built on by
Citing papers per year: 2024: 1; 2025: 1.
2 papers.