NoScope: Optimizing Neural Network Queries over Video at Scale

TL;DR

NoScope is a system that significantly accelerates neural network-based video queries by automatically constructing specialized, cost-effective model cascades tailored to specific videos and objects, reducing computational costs while maintaining high accuracy.

Contribution

It introduces an automated, inference-optimized model search and cascade construction approach that adapts to individual videos and objects, achieving large speed-ups over traditional neural network inference.

Findings

Achieves 265-15,500x real-time speed-up on binary classification tasks.

Maintains accuracy within 1-5% of state-of-the-art neural networks.

Reduces computational cost by up to three orders of magnitude.

Abstract

Recent advances in computer vision-in the form of deep neural networks-have made it possible to query increasing volumes of video data with high accuracy. However, neural network inference is computationally expensive at scale: applying a state-of-the-art object detector in real time (i.e., 30+ frames per second) to a single video requires a $4000 GPU. In response, we present NoScope, a system for querying videos that can reduce the cost of neural network video analysis by up to three orders of magnitude via inference-optimized model search. Given a target video, object to detect, and reference neural network, NoScope automatically searches for and trains a sequence, or cascade, of models that preserves the accuracy of the reference network but is specialized to the target video and are therefore far less computationally expensive. NoScope cascades two types of models: specialized models that forego the full generality of the reference model but faithfully mimic its behavior for the target video and object; and difference detectors that highlight temporal differences across frames. We show that the optimal cascade architecture differs across videos and objects, so NoScope uses an efficient cost-based optimizer to search across models and cascades. With this approach, NoScope achieves two to three order of magnitude speed-ups (265-15,500x real-time) on binary classification tasks over fixed-angle webcam and surveillance video while maintaining accuracy within 1-5% of state-of-the-art neural networks.