Borrowing from yourself: Faster future video segmentation with partial channel update

02/11/2022
by   Evann Courdier, et al.
0

Semantic segmentation is a well-addressed topic in the computer vision literature, but the design of fast and accurate video processing networks remains challenging. In addition, to run on embedded hardware, computer vision models often have to make compromises on accuracy to run at the required speed, so that a latency/accuracy trade-off is usually at the heart of these real-time systems' design. For the specific case of videos, models have the additional possibility to make use of computations made for previous frames to mitigate the accuracy loss while being real-time. In this work, we propose to tackle the task of fast future video segmentation prediction through the use of convolutional layers with time-dependent channel masking. This technique only updates a chosen subset of the feature maps at each time-step, bringing simultaneously less computation and latency, and allowing the network to leverage previously computed features. We apply this technique to several fast architectures and experimentally confirm its benefits for the future prediction subtask.

READ FULL TEXT

page 1

page 2

page 3

page 4

research
03/16/2021

GSVNet: Guided Spatially-Varying Convolution for Fast Semantic Segmentation on Video

This paper addresses fast semantic segmentation on video.Video segmentat...
research
04/03/2020

Temporally Distributed Networks for Fast Video Semantic Segmentation

We present TDNet, a temporally distributed network designed for fast and...
research
04/02/2018

Low-Latency Video Semantic Segmentation

Recent years have seen remarkable progress in semantic segmentation. Yet...
research
04/03/2020

Temporally Distributed Networks for Fast Video Segmentation

We present TDNet, a temporally distributed network designed for fast and...
research
08/03/2010

Fully automatic extraction of salient objects from videos in near real-time

Automatic video segmentation plays an important role in a wide range of ...
research
05/26/2020

CARPe Posterum: A Convolutional Approach for Real-time Pedestrian Path Prediction

Pedestrian path prediction is an essential topic in computer vision and ...
research
01/08/2013

Causal graph-based video segmentation

Numerous approaches in image processing and computer vision are making u...

Please sign up or login with your details

Forgot password? Click here to reset