ESWEEK 2021
who it's for
embedded machine learning engineers, edge AI researchers, hardware accelerator designers, FPGA developers, systems researchers, autonomous vehicle researchers, ML tooling engineers, neuromorphic engineers
topics
real-time edge video analytics; in-memory computation and RowClone; autonomous racing software stack; neural network and FPGA co-design; quantization-aware training with Brevitas; tinyML on microcontrollers; few-shot face verification; surrogate-gradient spiking training; memristor neuromorphic chips; FPGA streaming dataflow inference; reproducibility metrics for ML experiments; ML-driven manycore mapping
| title | author | length |
|---|---|---|
| Edge AI Systems | Lin Wang | 01h 56m |
Covers how to get real-time, reliable video analytics by moving work from cloud or device into an edge layer and gives concrete systems: Maui methods and… | ||
| Memory-Centric Computing | Onur Mutlu | 02h 00m |
Memory movement dominates energy and latency — measured workloads spend ~60% of device energy on data movement — so designs put simple computation into memory… | ||
| Learn to Drive (and Race!) Autonomous Vehicles | Rahul Mangharam, Johannes Betz | 01h 51m |
Hands-on course that puts working code on a 1/10-scale F1TENTH car (NVIDIA Jetson + LiDAR + camera) and in two simulators (f110_gym and Unity/LGSVL). Students… | ||
| Neural Networks and Accelerator Co-Design | Nicholas Fraser | 01h 55m |
Co-design of network and FPGA gives concrete gains: using Brevitas for quantization-aware training plus the FIN toolchain produces hardware-mapped networks… | ||
| TinyML: Machine Learning’s Future is Tiny and Bright | Vijay Janapa Reddi | 01h 53m |
Argues TinyML runs on-device at under 1 mW and on kilobytes of RAM, not megabytes, enabling always‑on sensor analytics; demonstrates a 2 KB runtime kernel and… | ||
| Face Verification using Few-shot Deep Learning | Amit Sethi | 01h 04m |
Deploys FaceNet on Jetson Nano using PyTorch for training and TensorRT to convert the model to FP16 with TensorRT for edge inference, with camera integration… | ||
| Spiking Neural Networks | Priya Panda | 01h 59m |
Survey of training methods, measured tradeoffs and working tools: surrogate‑gradient backprop appears viable and, when combined with Batch Normalization… | ||
| Neural Network Accelerator Design | Yu Wang | 01h 52m |
Survey of >200 papers and 80 products compares CPUs, GPUs, FPGAs and ASICs and shows FPGA <1 TOPS/W vs ASICs pushing 10s–100s TOPS/W; quantization and pruning… | ||
| Research Reproducibility in Embedded Learning | Romain Jacob | 01h 57m |
Introduces a reproducibility methodology implemented as the TriScale framework that uses nonparametric percentile confidence intervals (Thompson-style) to turn… | ||
| Introduction to Neuromorphic Computing | Hai Li, Ph.D. | 01h 57m |
Introduces a memristor-based neuromorphic chip with memristor crossbar arrays and a spiking processing engine that maps convolution kernels via a PipeLayer… | ||
| DNNs on FPGAs | Jae-sun Seo | 01h 54m |
Covers concrete FPGA strategies: loop unrolling, tiling and interchange with a worked case that maps about 3,136 PEs in an output‑stationary MAC array, plus… | ||
| Machine Learning for Manycore System Design and Optimization | Biresh Kumar Joardar, Jana Doppa | 01h 53m |
Survey of manycore basics and concrete problems: shows task mapping and network-on-chip as combinatorial searches whose evaluations need expensive… | ||
| nothing matches | ||