AI/ML Systems and Infrastructure SAND A New Programming Abstraction for Video-based Deep Learning SpecEdge A scalable edge-assisted framework that achieves enhanced cost efficiency and reduced latency StellaTrain A holistic framework that achieves near-optimal training speeds in multi-cloud environments. TopFull An Adaptive Top-Down Overload Control for Microservices. ES-MoE Efficient scaling of Mixture-of-Experts training on a limited number of GPUs. TSPipe Train knowledge distillation (KD) and self-supervised learning (SSL) faster with pipelines BWA-MEME BWA-MEM emulated with a machine learning approach GRAF A GNN-based proactive resource allocation framework for SLO-oriented microservices. AI-enabled Media Delivery NarrAD AI-based system for generating barrier-free audio descriptions AccelIR Accelerate image-restoration for image compression NeuroScaler Enable neural-enhanced video streaming at scale NEMO Run real-time video super-resolution on commercial smartphones LiveNAS Stream high-quality live videos even when the network becomes congested NAS Watch videos in high-quality even when the network becomes congested Intelligent Networked Systems/Systems Security FlexPass Flexible and practical proactive transport protocol for datacenter networks OutRAN A cell-scale flow scheduler designed for Radio Access Networks TLT Timeout-less Transport in Commodity Datacenter Networks APPx Automatically generates acceleration proxies for mobile apps TEE Systems Security-enhanced network system by leveraging trusted execution environment (TEE) DFC High-Performance String Matching Algorithm that accelerates network middlebox applications ExpressPass Credit-Scheduled Delay-Bounded Congestion Control for Datacenters
