TY - GEN
T1 - Iterative Pruning-based Model Compression for Pose Estimation on Resource-constrained Devices
AU - Choi, Sung Hyun
AU - Choi, Wonje
AU - Lee, Youngseok
AU - Woo, Honguk
N1 - Publisher Copyright:
© 2022 ACM.
PY - 2022/2/18
Y1 - 2022/2/18
N2 - In this work, we propose a pruning-based model compression scheme, aiming at achieving an efficient model that has strength in both accuracy and inference time on an embedded device environment with limited resources. The proposed scheme consists of (1) pruning profiling and (2) iterative pruning via knowledge distillation. With the scheme, we develop a resource-efficient 2D pose estimation model using HRNet and evaluate the model on NVIDA JetsonNano with the Microsoft COCO keypoint dataset. Specifically, our compressed model obtains the fast pose estimation of 20.3 FPS on NVIDA JetsonNano, while maintaining a high accuracy of 74.1 AP. Compared to the conventional HRNet model without compression, the proposed compression technique achieves 33 % improvement in FPS with only 0.4 % degradation in AP.
AB - In this work, we propose a pruning-based model compression scheme, aiming at achieving an efficient model that has strength in both accuracy and inference time on an embedded device environment with limited resources. The proposed scheme consists of (1) pruning profiling and (2) iterative pruning via knowledge distillation. With the scheme, we develop a resource-efficient 2D pose estimation model using HRNet and evaluate the model on NVIDA JetsonNano with the Microsoft COCO keypoint dataset. Specifically, our compressed model obtains the fast pose estimation of 20.3 FPS on NVIDA JetsonNano, while maintaining a high accuracy of 74.1 AP. Compared to the conventional HRNet model without compression, the proposed compression technique achieves 33 % improvement in FPS with only 0.4 % degradation in AP.
KW - Embedded System inference
KW - Knowledge Distillation
KW - Model compression
KW - Pose estimation
KW - Pruning
UR - https://www.scopus.com/pages/publications/85130321534
U2 - 10.1145/3523111.3523128
DO - 10.1145/3523111.3523128
M3 - Conference contribution
AN - SCOPUS:85130321534
T3 - ACM International Conference Proceeding Series
SP - 110
EP - 115
BT - ICMVA 2022 - 5th International Conference on Machine Vision and Applications
PB - Association for Computing Machinery
T2 - 5th International Conference on Machine Vision and Applications, ICMVA 2022
Y2 - 18 February 2022 through 20 February 2022
ER -