Sorry, you need to enable JavaScript to visit this website.

facebooktwittermailshare

On Intra Video Coding and In-Loop Filtering for Neural Object Detection Networks

Abstract: 

Classical video coding for satisfying humans as the final user is a widely investigated field of studies for visual content, and common video codecs are all optimized for the human visual system (HVS). But are the assumptions and optimizations also valid when the compressed video stream is analyzed by a machine? To answer this question, we compared the performance of two state-of-the-art neural detection networks when being fed with deteriorated input images coded with HEVC and VVC in an autonomous driving scenario using intra coding. Additionally, the impact of the three VVC in-loop filters when coding images for a neural network is examined. The results are compared using the mean average precision metric to evaluate the object detection performance for the compressed inputs. Throughout these tests, we found that the Bjøntegaard Delta Rate savings with respect to PSNR of 22.2 % using VVC instead of HEVC cannot be reached when coding for object detection networks with only 13.6 % in the best case. Besides, it is shown that disabling the VVC in-loop filters SAO and ALF results in bitrate savings of 6.4 % compared to the standard VTM at the same mean average precision.

up
0 users have voted:

Paper Details

Authors:
Kristian Fischer, Christian Herglotz, André Kaup
Submitted On:
9 November 2020 - 4:02am
Short Link:
Type:
Presentation Slides
Event:
Presenter's Name:
Kristian Fischer
Paper Code:
1436
Document Year:
2020
Cite

Document Files

presentation_ICIP_2020.pdf

(10)

Subscribe

[1] Kristian Fischer, Christian Herglotz, André Kaup, "On Intra Video Coding and In-Loop Filtering for Neural Object Detection Networks", IEEE SigPort, 2020. [Online]. Available: http://sigport.org/5549. Accessed: Nov. 29, 2020.
@article{5549-20,
url = {http://sigport.org/5549},
author = {Kristian Fischer; Christian Herglotz; André Kaup },
publisher = {IEEE SigPort},
title = {On Intra Video Coding and In-Loop Filtering for Neural Object Detection Networks},
year = {2020} }
TY - EJOUR
T1 - On Intra Video Coding and In-Loop Filtering for Neural Object Detection Networks
AU - Kristian Fischer; Christian Herglotz; André Kaup
PY - 2020
PB - IEEE SigPort
UR - http://sigport.org/5549
ER -
Kristian Fischer, Christian Herglotz, André Kaup. (2020). On Intra Video Coding and In-Loop Filtering for Neural Object Detection Networks. IEEE SigPort. http://sigport.org/5549
Kristian Fischer, Christian Herglotz, André Kaup, 2020. On Intra Video Coding and In-Loop Filtering for Neural Object Detection Networks. Available at: http://sigport.org/5549.
Kristian Fischer, Christian Herglotz, André Kaup. (2020). "On Intra Video Coding and In-Loop Filtering for Neural Object Detection Networks." Web.
1. Kristian Fischer, Christian Herglotz, André Kaup. On Intra Video Coding and In-Loop Filtering for Neural Object Detection Networks [Internet]. IEEE SigPort; 2020. Available from : http://sigport.org/5549