08:15 AM Opening (Chairs)

08:30 AM Aurélien Bellet (Contributed talk)

          • Communication-Efficient Decentralized Boosting while Discovering the Collaboration Graph

08:45 AM Neel Guha (Contributed talk)

          • One-Shot Federated Learning

09:00 AM Prof. Kurt Keutzer (Invited talk)

          • The Search for Mobile Models

09:30 AM Ting-Wu Chin (Contributed talk)

          • Layer-compensated Pruning for Resource-constrained Convolutional Neural Networks

09:45 AM Prof. Thad Starner (Keynote talk)

          • Symbiotic AI: Using wearable devices to teach computers how to live in a human world

10:30 AM Coffee break (morning) (Break)

11:00 AM Prof. Max Welling (Invited talk)

          • The Deep Integration of Hardware and Machine Learning; with an application to hardware-aware quantization

11:30 AM Zornitsa Kozareva (Contributed talk)

          • Fast & Small On-device Neural Networks for Short Text Natural Language Processing

11:50 AM Spotlight (poster, demo), Lunch & Poster Session (Spotlight & Poster)

Brijraj Singh, phildow Dow, Robert Dürichen, Paul Whatmough, Chen Feng, Arijit Patra, Shishir Patil, Eunjeong Jeong, Desmond Lin, Yuki Izumi, Isabelle Leang, Mimee Xu, wenhan zhang, Sam Witteveen

01:30 PM Brendan McMahan (Invited talk)

          • Federated Learning in Practice at Google

02:00 PM Prof. Virginia Smith (Invited talk)

          • Learning in Heterogeneous Networks

02:30 PM Meghan Cowan (Contributed talk)

          • Towards Automating Generation of Low Precision Deep Learning Operators

02:45 PM Kuan Wang (Contributed talk)

          • Learning to Quantize Neural Networks

03:00 PM Coffee break (afternoon) (Break)

03:30 PM Jan Kautz (Invited talk)

          • Efficient Computer Vision for Intelligent User Interfaces

04:00 PM Prof. Song Han (Invited talk)

          • Proxyless Architecture Search for Target Task and Hardware: from Images to Videos

04:30 PM Demo session (Demo)

Sonam Damani, phildow Dow, Yuki Izumi, Shishir Patil, Isabelle Leang, Mimee Xu, wenhan zhang