Towards the on-device Handwriting Trajectory Reconstruction of the Sensor Enhanced Pen - Inria - Institut national de recherche en sciences et technologies du numérique Accéder directement au contenu
Communication Dans Un Congrès Année : 2023

Towards the on-device Handwriting Trajectory Reconstruction of the Sensor Enhanced Pen

Résumé

Performing handwriting trajectory regression from inertial data using Deep Neural Network (DNN) on an embedded device is a very challenging task, since the network accuracy is prone to imperfections in the weights and needs a significant amount of parameters to be able to regress. In this work, we apply and compare different quantization techniques and Mitchell logarithmic multiplication approximation in order to enable the on-device inference. We show that it is possible to perform the inference of the TCN-based regression model using only 8-bit fixed-point quantization without significant reconstruction precision loss and that the accuracy degradation of the approximate multiplication can be partially compensated with Quantizationaware Training (QAT). Finally, we demonstrate that the compressed models can be integrated into an off-the-shelf commercial Systems-on-Chip with minimal use of FPU and requiring only 460 KB of the ROM size for the TCN-49 configuration.
Fichier principal
Vignette du fichier
WFIoT_2023_KIHT_joint_paper.pdf (559.67 Ko) Télécharger le fichier
Origine : Fichiers produits par l'(les) auteur(s)

Dates et versions

hal-04358219 , version 1 (21-12-2023)

Licence

Paternité

Identifiants

  • HAL Id : hal-04358219 , version 1

Citer

Alexey Serdyuk, Fabian Kreß, Micha Hiegle, Tanja Harbaum, Jürgen Becker, et al.. Towards the on-device Handwriting Trajectory Reconstruction of the Sensor Enhanced Pen. IEEE 9th World Forum on Internet of Things, Oct 2023, Aveiro, Portugal. ⟨hal-04358219⟩
52 Consultations
21 Téléchargements

Partager

Gmail Facebook X LinkedIn More