2019 IEEE Computer Society Annual Symposium on VLSI (ISVLSI) 2019
DOI: 10.1109/isvlsi.2019.00012
|View full text |Cite
|
Sign up to set email alerts
|

T-DLA: An Open-source Deep Learning Accelerator for Ternarized DNN Models on Embedded FPGA

Help me understand this report

Search citation statements

Order By: Relevance

Paper Sections

Select...
2
1
1
1

Citation Types

0
35
0

Year Published

2020
2020
2024
2024

Publication Types

Select...
4
3

Relationship

1
6

Authors

Journals

citations
Cited by 34 publications
(35 citation statements)
references
References 10 publications
0
35
0
Order By: Relevance
“…Even with the k-means based solutions, the distribution of the weights in the same layer is ignored in the quantization process. However, the consideration of the distribution of the weight data is proven to be effective for the accuracy control in the existing approaches [8], [18].…”
Section: Motivation Of the Vecq Methodsmentioning
confidence: 99%
See 3 more Smart Citations
“…Even with the k-means based solutions, the distribution of the weights in the same layer is ignored in the quantization process. However, the consideration of the distribution of the weight data is proven to be effective for the accuracy control in the existing approaches [8], [18].…”
Section: Motivation Of the Vecq Methodsmentioning
confidence: 99%
“…As an effective way to compress DNNs, many quantization methods have been explored [6], [8], [9], [10], [11], [12], [13], [14], [15], [16], [17], [18], [19], [20], [21], [22], [23], [24], [25], [26], [27], [28]. These quantization methods can be roughly categorized into 3 different types based on their objective functions for the quantization process:…”
Section: Related Work and Motivationmentioning
confidence: 99%
See 2 more Smart Citations
“…Finally, a recent trend in embedded machine learning for IoT devices includes the use of hardware accelerators for neural networks. Examples can be found in academic research [26][27][28] and off-the-shelf industrial solutions [29]. Although leveraging such hardware is important whenever available, our work targets wearable systems with no special hardware capabilities; hence, the proposed framework operates on general-purpose microcontrollers and is backwards-compatible to legacy IoT systems.…”
Section: Related Workmentioning
confidence: 99%