Conference paper Open Access

Pruned Lightweight Encoders for Computer Vision

Jakub Žádník; Markku Mäkitalo; Pekka Jääskeläinen

Latency-critical computer vision systems, such as autonomous driving or drone control, require fast image or video compression when offloading neural network inference to a remote computer. To ensure low latency on a near-sensor edge device, we propose the use of lightweight encoders with constant bitrate and pruned encoding configurations, namely, ASTC and JPEG XS. Pruning introduces significant distortion which we show can be recovered by retraining the neural network with compressed data after decompression. Such an approach does not modify the network architecture or require coding format modifications. By retraining with compressed datasets, we reduced the classification accuracy and segmentation mean intersection over union (mIoU) degradation due to ASTC compression to 4.9–5.0 percentage points (pp) and 4.4–4.0 pp, respectively. With the same method, the mIoU lost due to JPEG XS compression at the main profile was restored to 2.7–2.3 pp. In terms of encoding speed, our ASTC encoder implementation is 2.3x faster than JPEG. Even though the JPEG XS reference encoder requires optimizations to reach low latency, we showed that disabling significance flag coding saves 22–23% of encoding time at the cost of 0.4–0.3 mIoU after retraining.

Files (314.5 kB)
Name Size
Pruned_Lightweight_Encoders_for_Computer_Vision__MMSP_2022_.pdf
md5:b0af178699e384aefa6141e8ac37e09e
314.5 kB Download
33
24
views
downloads
All versions This version
Views 3333
Downloads 2424
Data volume 7.5 MB7.5 MB
Unique views 2929
Unique downloads 2424

Share

Cite as