Goto

Collaborating Authors

 synchronization








High-ThroughputSynchronousDeepRL

Neural Information Processing Systems

Deep reinforcement learning (RL) is computationally demanding and requiresprocessing of many data points. Synchronous methods enjoy training stability while having lowerdatathroughput.




LearningRepresentationsfromAudio-Visual SpatialAlignment

Neural Information Processing Systems

While these approaches learn high-quality representations for downstream tasks such as action recognition, their training objectives disregard spatial cues naturally occurring in audio and visual signals.