TY - GEN
T1 - Performance of Texture Compression Algorithms in Low-Latency Computer Vision Tasks
AU - Zadnik, Jakub
AU - Mäkitalo, Markku
AU - Iho, Jussi
AU - Jääskeläinen, Pekka
N1 - JUFOID=71968
Funding Information:
The work was financially supported by the Tampere University ITC Graduate School. This project has received funding from the ECSEL Joint Undertaking (JU) under grant agreement No 783162. The JU receives support from the European Union’s Horizon 2020 research and innovation programme and Netherlands, Czech Republic, Finland, Spain, Italy. It was also supported by European Union’s Horizon 2020 research and innovation programme under Grant Agreement No 871738 (CPSoSAware).
Publisher Copyright:
© 2021 IEEE.
PY - 2021
Y1 - 2021
N2 - Deep learning has been successfully used for computer vision tasks, but its high computational cost limits the adoption in lightweight devices such as camera sensors. For this reason, many low-latency vision systems offload the inference computation to a local server, requiring fast (de)compression of the source images. Texture compression is a compelling alternative to existing compression schemes, such as JPEG or HEVC, due to its low decoding overhead, straightforward parallelization, robustness, and a fixed compression ratio. In this paper, we study the impact of lightweight bounding box-based texture compression algorithms, BC1 and YCoCg-BC3, on the accuracy of two computer vision tasks: object detection and semantic segmentation. While JPEG achieves superior per-pixel error rate, the YCoCg-BC3 encoding can provide comparable vision accuracy. The BC1 encoding results in significant degradation of vision performance. However, by retraining the FasterSeg teacher network with a BC1-compressed dataset, we reduced its segmentation mIoU loss from 2.7 to 0.5 percent. Thus, both BC1 and YCoCg-BC3 encoders are suitable for use in low latency vision systems, since they both achieve significantly higher encoding speed than JPEG and their decoding overhead is negligible.
AB - Deep learning has been successfully used for computer vision tasks, but its high computational cost limits the adoption in lightweight devices such as camera sensors. For this reason, many low-latency vision systems offload the inference computation to a local server, requiring fast (de)compression of the source images. Texture compression is a compelling alternative to existing compression schemes, such as JPEG or HEVC, due to its low decoding overhead, straightforward parallelization, robustness, and a fixed compression ratio. In this paper, we study the impact of lightweight bounding box-based texture compression algorithms, BC1 and YCoCg-BC3, on the accuracy of two computer vision tasks: object detection and semantic segmentation. While JPEG achieves superior per-pixel error rate, the YCoCg-BC3 encoding can provide comparable vision accuracy. The BC1 encoding results in significant degradation of vision performance. However, by retraining the FasterSeg teacher network with a BC1-compressed dataset, we reduced its segmentation mIoU loss from 2.7 to 0.5 percent. Thus, both BC1 and YCoCg-BC3 encoders are suitable for use in low latency vision systems, since they both achieve significantly higher encoding speed than JPEG and their decoding overhead is negligible.
KW - Computer Vision
KW - Image Compression
KW - Low Latency
KW - Texture Compression
U2 - 10.1109/EUVIP50544.2021.9484015
DO - 10.1109/EUVIP50544.2021.9484015
M3 - Conference contribution
AN - SCOPUS:85111442988
SN - 9781665432313
T3 - European Workshop on Visual Information Processing
BT - Proceedings of the 2021 9th European Workshop on Visual Information Processing, EUVIP 2021
A2 - Beghdadi, A.
A2 - Cheikh, F. Alaya
A2 - Tavares, J.M.R.S.
A2 - Mokraoui, A.
A2 - Valenzise, G.
A2 - Oudre, L.
A2 - Qureshi, M.A.
PB - IEEE
T2 - European Workshop on Visual Information Processing
Y2 - 23 June 2021 through 25 June 2021
ER -