Mobile devices such as drones and autonomous vehicles increasingly rely on object detection (OD) through deep neural networks (DNNs) to perform critical tasks such as navigation, target-tracking and surveillance, just to name a few. Due to their high complexity, the execution of these DNNs requires excessive time and energy. Low-complexity object tracking (OT) is thus used along with OD, where the latter is periodically applied to generate "fresh" references for tracking. However, the frames processed with OD incur large delays, which does not comply with real-time applications requirements. Offloading OD to edge servers can mitigate this issue, but existing work focuses on the optimization of the offloading process in systems where the wireless channel has a very large capacity. Herein, we consider systems with constrained and erratic channel capacity, and establish parallel OT (at the mobile device) and OD (at the edge server) processes that are resilient to large OD latency. We propose Katch-Up, a novel tracking mechanism that improves the system resilience to excessive OD delay. We show that this technique greatly improves the quality of the reference available to tracking, and boosts performance up to 33%. However, while Katch-Up significantly improves performance, it also increases the computing load of the mobile device. Hence, we design SmartDet, a low-complexity controller based on deep reinforcement learning (DRL) that learns to achieve the right trade-off between resource utilization and OD performance. SmartDet takes as input highly-heterogeneous context-related information related to the current video content and the current network conditions to optimize frequency and type of OD offloading, as well as Katch-Up utilization. We extensively evaluate SmartDet on a real-world testbed composed by a JetSon Nano as mobile device and a GTX 980 Ti as edge server, connected through a Wi-Fi link, to collect several network-related traces, as well as energy measurements. We consider a state-of-the-art video dataset (ILSVRC 2015 - VID) and state-of-the-art OD models (EfficientDet 0, 2 and 4). Experimental results show that SmartDet achieves an optimal balance between tracking performance – mean Average Recall (mAR) and resource usage. With respect to a baseline with full Katch-Up usage and maximum channel usage, we still increase mAR by 4% while using 50% less of the channel and 30% power resources associated with Katch-Up. With respect to a fixed strategy using minimal resources, we increase mAR by 20% while using Katch-Up on 1/3 of the frames.
more »
« less
Do we need sophisticated system design for edge-assisted augmented reality?
We revisit the performance of a canonical system design for edge-assisted AR that simply combines off-the-shelf H.264 video encoding with a standard object tracking technique. Our experimental analysis shows that the simple canonical design for edge-assisted object detection can achieve within 3.07%/1.51% of the accuracy of ideal offloading (which assumes infinite network bandwidth and the total network transmission time of a single RTT) under LTE/5G mmWave networks. Our findings suggest that recent trend towards sophisticated system architecture design for edge-assisted AR appears unnecessary. We provide insights for why video compression plus on-device object tracking is so effective in edge-assisted object detection, draw implications to edge-assisted AR research, and pose open problems that warrant further investigation into this surprise finding.
more »
« less
- Award ID(s):
- 2112778
- NSF-PAR ID:
- 10342971
- Date Published:
- Journal Name:
- EdgeSys '22: Proceedings of the 5th International Workshop on Edge Systems, Analytics and Networking
- Page Range / eLocation ID:
- 7 to 12
- Format(s):
- Medium: X
- Sponsoring Org:
- National Science Foundation
More Like this
-
-
Recent breakthroughs in deep learning (DL) have led to the emergence of many intelligent mobile applications and services, but in the meanwhile also pose unprecedented computing challenges on resource-constrained mobile devices. This paper builds a collaborative deep inference system between a resource-constrained mobile device and a powerful edge server, aiming at joining the power of both on-device processing and computation offloading. The basic idea of this system is to partition a deep neural network (DNN) into a front-end part running on the mobile device and a back-end part running on the edge server, with the key challenge being how to locate the optimal partition point to minimize the end-to-end inference delay. Unlike existing efforts on DNN partitioning that rely heavily on a dedicated offline profiling stage to search for the optimal partition point, our system has a built-in online learning module, called Autodidactic Neurosurgeon (ANS), to automatically learn the optimal partition point on-the-fly. Therefore, ANS is able to closely follow the changes of the system environment by generating new knowledge for adaptive decision making. The core of ANS is a novel contextual bandit learning algorithm, called μLinUCB, which not only has provable theoretical learning performance guarantee but also is ultra-lightweight for easy real-world implementation. We implement our system on a video stream object detection testbed to validate the design of ANS and evaluate its performance. The experiments show that ANS significantly outperforms state-of-the-art benchmarks in terms of tracking system changes and reducing the end-to-end inference delay.more » « less
-
The traffic congestion hits most big cities in the world - threatening long delays and serious reductions in air quality. City and local government officials continue to face challenges in optimizing crowd flow, synchronizing traffic and mitigating threats or dangerous situations. One of the major challenges faced by city planners and traffic engineers is developing a robust traffic controller that eliminates traffic congestion and imbalanced traffic flow at intersections. Ensuring that traffic moves smoothly and minimizing the waiting time in intersections requires automated vehicle detection techniques for controlling the traffic light automatically, which are still challenging problems. In this paper, we propose an intelligent traffic pattern collection and analysis model, named TPCAM, based on traffic cameras to help in smooth vehicular movement on junctions and set to reduce the traffic congestion. Our traffic detection and pattern analysis model aims at detecting and calculating the traffic flux of vehicles and pedestrians at intersections in real-time. Our system can utilize one camera to capture all the traffic flows in one intersection instead of multiple cameras, which will reduce the infrastructure requirement and potential for easy deployment. We propose a new deep learning model based on YOLOv2 and adapt the model for the traffic detection scenarios. To reduce the network burdens and eliminate the deployment of network backbone at the intersections, we propose to process the traffic video data at the network edge without transmitting the big data back to the cloud. To improve the processing frame rate at the edge, we further propose deep object tracking algorithm leveraging adaptive multi-modal models and make it robust to object occlusions and varying lighting conditions. Based on the deep learning based detection and tracking, we can achieve pseudo-30FPS via adaptive key frame selection.more » « less
-
Edge-assisted Augmented Reality (AR) which offloads computeintensive Deep Neural Network (DNN)-based AR tasks to edge servers faces an important design challenge: how to pick the DNN model out of many choices proposed for each AR task for offloading. For each AR task, e.g., depth estimation, many DNN-based models have been proposed over time that vary in accuracy and complexity. In general, more accurate models are also more complex; they are larger and have longer inference time. Thus choosing a larger model in offloading can provide higher accuracy for the offloaded frames but also incur longer turnaround time, during which the AR app has to reuse the estimation result from the last offloaded frame, which can lead to lower average accuracy. In this paper, we experimentally study this design tradeoff using depth estimation as a case study. We design optimal offloading schedule and further consider the impact of numerous factors such as on-device fast tracking, frame downsizing and available network bandwidth. Our results show that for edge-assisted monocular depth estimation, with proper frame downsizing and fast tracking, compared to small models, the improved accuracy of large models can offset its longer turnaround time to provide higher average estimation accuracy across frames under both LTE and 5G mmWave.more » « less
-
Mobile Augmented Reality (AR), which overlays digital content on the real-world scenes surrounding a user, is bringing immersive interactive experiences where the real and virtual worlds are tightly coupled. To enable seamless and precise AR experiences, an image recognition system that can accurately recognize the object in the camera view with low system latency is required. However, due to the pervasiveness and severity of image distortions, an effective and robust image recognition solution for “in the wild” mobile AR is still elusive. In this article, we present CollabAR, an edge-assisted system that provides distortion-tolerant image recognition for mobile AR with imperceptible system latency. CollabAR incorporates both distortion-tolerant and collaborative image recognition modules in its design. The former enables distortion-adaptive image recognition to improve the robustness against image distortions, while the latter exploits the spatial-temporal correlation among mobile AR users to improve recognition accuracy. Moreover, as it is difficult to collect a large-scale image distortion dataset, we propose a Cycle-Consistent Generative Adversarial Network-based data augmentation method to synthesize realistic image distortion. Our evaluation demonstrates that CollabAR achieves over 85% recognition accuracy for “in the wild” images with severe distortions, while reducing the end-to-end system latency to as low as 18.2 ms.more » « less
- Free Publicly Accessible Full Text
-
Accepted Manuscript1.0
- Conference Paper:
- https://doi.org/10.1145/3517206.3526267
