PORTO @ 
Archivio Istituzionale della Ricerca
IRIS è la soluzione IT che facilita la raccolta e la gestione dei dati relativi alle attività e ai prodotti della ricerca. Fornisce a ricercatori, amministratori e valutatori gli strumenti per monitorare i risultati della ricerca, aumentarne la visibilità e allocare in modo efficace le risorse disponibili.
EnglishConvolutional Neural Networks (ConvNets) are trained offline using the few available data and may therefore suffer from substantial accuracy loss when ported on the field, where unseen input patterns received under unpredictable external conditions can mislead the model. Test-Time Augmentation (TTA) techniques aim to alleviate such common side effect at inference-time, first running multiple feed-forward passes on a set of altered versions of the same input sample, and then computing the main outcome through a consensus of the aggregated predictions. Unfortunately, the implementation of TTA on embedded CPUs introduces latency penalties that limit its adoption on edge applications. To tackle this issue, we propose AdapTTA, an adaptive implementation of TTA that controls the number of feed-forward passes dynamically, depending on the complexity of the input. Experimental results on state-of-the-art ConvNets for image classification deployed on a commercial ARM Cortex-A CPU demonstrate AdapTTA reaches remarkable latency savings, from 1.40× to 2.21×, and hence a higher frame rate compared to static TTA, still preserving the same accuracy gain.
AdapTTA: Adaptive Test-Time Augmentation for Reliable Embedded ConvNets / Mocerino, Luca; Rizzo, Roberto G.; Peluso, Valentino; Calimera, Andrea; Macii, Enrico. - (2021), pp. 1-6. ((Intervento presentato al convegno International Conference on Very Large Scale Integration (VLSI-SoC) nel 4-7 Oct. 2021 [10.1109/VLSI-SoC53125.2021.9606980].
| Titolo: | AdapTTA: Adaptive Test-Time Augmentation for Reliable Embedded ConvNets | |
| Autori: | ||
| Data di pubblicazione: | 2021 | |
| Abstract: | Convolutional Neural Networks (ConvNets) are trained offline using the few available data and may... therefore suffer from substantial accuracy loss when ported on the field, where unseen input patterns received under unpredictable external conditions can mislead the model. Test-Time Augmentation (TTA) techniques aim to alleviate such common side effect at inference-time, first running multiple feed-forward passes on a set of altered versions of the same input sample, and then computing the main outcome through a consensus of the aggregated predictions. Unfortunately, the implementation of TTA on embedded CPUs introduces latency penalties that limit its adoption on edge applications. To tackle this issue, we propose AdapTTA, an adaptive implementation of TTA that controls the number of feed-forward passes dynamically, depending on the complexity of the input. Experimental results on state-of-the-art ConvNets for image classification deployed on a commercial ARM Cortex-A CPU demonstrate AdapTTA reaches remarkable latency savings, from 1.40× to 2.21×, and hence a higher frame rate compared to static TTA, still preserving the same accuracy gain. | |
| ISBN: | 978-1-6654-2614-5 | |
| Appare nelle tipologie: | 4.1 Contributo in Atti di convegno |
File in questo prodotto:
| File | Descrizione | Tipologia | Licenza | |
|---|---|---|---|---|
| Adaptive_Test_time_Augmentation____VLSI_SoC2021.pdf | 2. Post-print / Author's Accepted Manuscript | PUBBLICO - Tutti i diritti riservati | Visibile a tuttiVisualizza/Apri | |
| AdapTTA_Adaptive_Test-Time_Augmentation_for_Reliable_Embedded_ConvNets.pdf | 2a Post-print versione editoriale / Version of Record | Non Pubblico - Accesso privato/ristretto | Administrator Richiedi una copia |
http://hdl.handle.net/11583/2947632
Copyright © 2022