Search

Video analytics pipelines incorporate on-premise edge servers to lower analysis latency, ensure privacy, and reduce bandwidth requirements. However, compared to the cloud, edge servers typically have lower processing power and GPU memory, limiting the number of video streams that they can manage and analyze. Existing solutions for memory management, such as swapping models in and out of GPU, having a common model stem, or compression and quantization to reduce the model size incur high overheads and often provide limited benefits. In this paper, we propose model merging as an approach towards memory management at the edge. This proposal is based on our observation that models at the edge share common layers, and that merging these common layers across models can result in significant memory savings. Our preliminary evaluation indicates that such an approach could result in up to 75% savings in the memory requirements. We conclude by discussing several challenges involved with realizing the model merging vision.

To cope with the high resource (network and compute) demands of real-time video analytics pipelines, recent systems have relied on frame filtering. However, filtering has typically been done with neural networks running on edge/backend servers that are expensive to operate. This paper investigates on-camera filtering, which moves filtering to the beginning of the pipeline. Unfortunately, we find that commodity cameras have limited compute resources that only permit filtering via frame differencing based on low-level video features. Used incorrectly, such techniques can lead to unacceptable drops in query accuracy. To overcome this, we built Reducto, a system that dynamically adapts filtering decisions according to the time-varying correlation between feature type, filtering threshold, query accuracy, and video content. Experiments with a variety of videos and queries show that Reducto achieves significant (51-97% of frames) filtering benefits, while consistently meeting the desired accuracy.

Named Data Networking (NDN) proposes a fundamental architectural change to the Internet, moving from point-to-point communication to a data-centric model. NDN-enabled nodes can communicate over any substrate that can deliver datagrams, such as layer-2 links (WiFi, BLE, Ethernet, etc.) and IP/UDP/TCP tunnels over IP connectivity. However in the latter case, NDN-enabled nodes must be able to discover the presence of each other and the data each serves in an automated way. This poster describes the design of an NDN Neighbor Discovery service (NDND), which enables isolated NDN nodes to discover each other and interconnect through tunneling over IP connectivity.