On the Efficient Representation and Execution of Deep Acoustic Models
The DOI matches its Crossref record; 4 of the 4 references it cites were found in a registry or an index.
Abstract
In this paper we present a simple and computationally efficient quantization scheme that enables us to reduce the resolution of the parameters of a neural network from 32-bit floating point values to 8-bit integer values.The proposed quantization scheme leads to significant memory savings and enables the use of optimized hardware instructions for integer arithmetic, thus significantly reducing the cost of inference.Finally, we propose a 'quantization aware' training process that applies the proposed scheme during network training and find that it allows us to recover most of the loss in accuracy introduced by quantization.We validate the proposed techniques by applying them to a long short-term memory-based acoustic model on an open-ended large vocabulary speech recognition task.
Overview
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Integrity signals
Registry
Cites
Statistics
Verification
Claims the paper makes
The claims and statistics of this paper haven’t been checked yet.
Free: it reads the open-access text and checks the first 5 claims that cite a source we can check. Create a free account to check every claim.
Self-check
Percentages, sums and reported differences re-computed from the paper’s own numbers.
The self-check needs the paper’s full text.
Citations
Cites
Graves, A., Fernández, S., Gomez, F., & Schmidhuber, J. (2006) · Proceedings of the 23rd international conference on Machine learning - ICML '06 · doi:10.1145/1143844.1143891 · Cited by 5,370
Deep Learning with Limited Numerical Precision
Gupta, S., Agrawal, A., Gopalakrishnan, K., & Narayanan, P. (2015) · arXiv · doi:10.48550/arxiv.1502.02551 · Cited by 1,364 · cited 1×
“Previous work has explored reducing the resolution even further [11], up to single bit representations [12, 13, 14, 15, 16, 17], and extending the types of topologies from fully-connected to convolutional neural networks (CNNs) [11, 15, 16, 18].”
BackgroundKim, M., & Smaragdis, P. (2016) · arXiv · doi:10.48550/arxiv.1601.06071 · Cited by 166 · cited 4×
“1, we make it part of the training process, under the principle that quantization noise must be considered when computing the model’s overall error, and thus gradients [14].”
BackgroundMethodInfluentialResiliency of Deep Neural Networks under Quantization
Sung, W., Shin, S., & Hwang, K. (2015) · arXiv · doi:10.48550/arxiv.1511.06488 · Cited by 132
Cited by
Streaming End-to-end Speech Recognition for Mobile Devices
He, Y., Sainath, T. N., Prabhavalkar, R., McGraw, I., Alvarez, R., Zhao, D., Rybach, D., Kannan, A., Wu, Y., Pang, R., et al. (2019) · ICASSP 2019 - 2019 IEEE International Conference on Acoustics, Speech and Signal Processing (ICASSP) · doi:10.1109/icassp.2019.8682336 · Cited by 601
Self-citationData-Free Quantization Through Weight Equalization and Bias Correction
Nagel, M., Baalen, M. V., Blankevoort, T., & Welling, M. (2019) · 2019 IEEE/CVF International Conference on Computer Vision (ICCV) · doi:10.1109/iccv.2019.00141 · Cited by 500
Edge Intelligence: Empowering Intelligence to the Edge of Network
Xu, D., Li, T., Li, Y., Su, X., Tarkoma, S., Jiang, T., Crowcroft, J., & Hui, P. (2021) · Proceedings of the IEEE · doi:10.1109/jproc.2021.3119950 · Cited by 272
Amortized Neural Networks for Low-Latency Speech Recognition
Macoskey, J., Strimel, G. P., Su, J., & Rastrow, A. (2021) · Interspeech 2021 · doi:10.21437/interspeech.2021-712 · Cited by 12
Metadata
- Authors
- Alvarez, R., Prabhavalkar, R., & Bakhtin, A.
- Venue
- Interspeech 2016
- Publisher
- ISCA
- Published
- 2016-09-08
- Pages
- 2746-2750
- ISSN
- 2308-457X (ISSN-L)
- OpenAlex
- W2497335382
- Cited by
- 44
- References
- 4
- FWCI
- 2.18
- Type
- conference paper
- Language
- en
- Open access
- closed
About these checks
Registry, retraction, Cites and statistics results are computed, not judged, with one exception: a printed reference that no index lists is read once by a model before it counts as a citation error. Claims are extracted by a model and kept only when their quote appears verbatim in the text; each cited source is then checked, with a model’s help, for whether it says what the claim relies on. Those verdicts can be wrong — read the claim’s quote and the cited source before relying on one.
Statistics and arithmetic are recomputed, not judged.
Confidence bands are shown only for fields calibrated on our annotated test set. None is calibrated yet, so no band is shown.
What counts as a claim ›How we measure confidence ›How we check this ›
Are you an author, or do you see a problem with this page? Request a review
Data sources
Bibliographic data: Semantic Scholar, OpenAlex, Crossref.
Contains information from the Semantic Scholar Academic Graph Datasets, made available under the ODC Attribution License (ODC-BY 1.0).
Semantic Scholar’s data is described in: Kinney, R., et al. (2023). The Semantic Scholar Open Data Platform. arXiv:2301.10140.