How to make on-device speech recognition practical

Branching encoder networks make operation more efficient, while “neural diffing” reduces bandwidth requirements for model updates.

Historically, Alexa’s automatic-speech-recognition models, which convert speech to text, have run in the cloud. But in recent years, we’ve been working to move more of Alexa’s computational capacity to the edge of the network — to Alexa-enabled devices themselves.

The move to the edge promises faster response times, since data doesn’t have to travel to and from the cloud; lower consumption of Internet bandwidth, which is important in some applications; and availability on devices with inconsistent Internet connections, such as Alexa-enabled in-car sound systems.

At this year’s Interspeech, we and our colleagues presented two papers describing some of the innovations we’re introducing to make it practical to run Alexa at the edge.

In one paper, “Amortized neural networks for low-latency speech recognition”, we show how to reduce the computational cost of neural-network-based automatic speech recognition (ASR) by 45% with no loss in accuracy. Our method also has lower latencies than similar methods for reducing computation, meaning that it enables Alexa to respond more quickly to customer requests.

In the other paper, “Learning a neural diff for speech models”, we show how to dramatically reduce the bandwidth required to update neural models on the edge. Instead of transmitting a complete model, we transmit a set of updates for some select parameters. In our experiments, this reduced the size of the update by as much as 98% with negligible effect on model accuracy.

Amortized neural networks

Neural ASR models are usually encoder-decoder models. The input to the encoder is a sequence of short speech snippets called frames, which the encoder converts into a representation that’s useful for decoding. The decoder translates that representation into text.

Neural encoders can be massive, requiring millions of computations for each input. But much of a speech signal is uninformative, consisting of pauses between syllables or redundant sounds. Passing uninformative frames through a huge encoder is just wasted computation.

Our approach is to use multiple encoders, of differing complexity, and decide on the fly which should handle a given frame of speech. That decision is made by a small neural network called an arbitrator, which must process every input frame before it’s encoded. The arbitrator adds some computational overhead to the process, but the time savings from using a leaner encoder is more than enough to offset it.

Researchers have tried similar approaches in domains other than speech, but when they trained their models, they minimized the average complexity of the frame-encoding process. That leaves open the possibility that the last few frames of the signal may pass to the more complex encoder, causing delays (increasing latency).

amortized-loss-2.png
Both processing flows above (a and b) distribute the same number of frames to the fast and slow (F and S) encoders, respectively, resulting in the same average computational cost. But the top flow incurs a significantly higher latency.

In our paper, we propose a new loss function that adds a penalty (Lamr in the figure above) for routing frames to the fast encoder when we don’t have a significant audio backlog. Without the penalty term, our branched-encoder model reduces latency to 29 to 234 milliseconds, versus thousands of milliseconds for models with a single encoder. But adding the penalty term cuts latency even further, to the 2-to-9-millisecond range.

AmazonScience_AmnetDemo_V1.gif
The audio backlog is one of the factors that the arbitrator considers when deciding which encoder should receive a given frame of audio.

In our experiments, we used two encoders, one complex and one lean, although in principle, our approach could generalize to larger numbers of encoders.

We train the arbitrator and both encoders together, end to end. During training, the same input passes through both encoders, and based on the accuracy of the resulting speech transcription, the arbitrator learns a probability distribution, which describes how often it should route frames with certain characteristics to the slow or fast encoder.

Over multiple epochs — multiple passes through the training data — we turn up the “temperature” on the arbitrator, skewing the distribution it learns more dramatically. In the first epoch, the split for a certain type of frame might be 70%-30% toward one encoder or the other. After three or four epochs, however, all of the splits are more like 99.99%-0.01% — essentially binary classifications.

We used three baselines in our experiments, all of which were single-encoder models. One was the full-parameter model, and the other two were compressed versions of the same model. One of these was compressed through sparsification (pruning of nonessential network weights), the other through matrix factorization (decomposing the model’s weight matrix into two smaller matrices that are multiplied together). 

Against the baselines, we compared two versions of our model, which were compressed through the same two methods. We ran all the models on a single-threaded processor at 650 million FLOPs per second.

Our sparse model had the lowest latency —two milliseconds, compared to 3,410 to 6,154 milliseconds for the baselines — and our matrix factorization model required the fewest number of floating-point operations per frame — 23 million, versus 30 million to 43 million for the baselines. Our accuracy remained comparable, however — a word error rate of 8.6% to 8.7%, versus 8.5% to 8.7% for the baselines.

Neural diffs

The ASR models that power Alexa are constantly being updated. During the Olympics, for instance, we anticipated a large spike in requests that used words like “Ledecky” and “Kalisz” and updated our models accordingly.

With cloud-based ASR, when we’ve updated a model, we simply send copies of it to a handful of servers in a data center. But with edge ASR, we may ultimately need to send updates to millions of devices simultaneously. So one of our research goals is to minimize the bandwidth requirements for edge updates.

In our other Interspeech paper, we borrow an idea from software engineering — that of the diff, or a file that charts the differences between the previous version of a codebase and the current one.

Our idea was that, if we could develop the equivalent of a diff for neural networks, we could use it to update on-device ASR models, rather than having to transmit all the parameters of a complete network with every update.

We experimented with two different approaches to creating a diff, matrix sparsification and hashing. With matrix sparsification we begin with two matrices of the same size, one that represents the weights of the connections in the existing ASR model and one that’s all zeroes.

Then, when we retrain the ASR model on new data, we update, not the parameters of the old model, but the entries in the second matrix — the diff. The updated model is a linear combination of the original weights and the values in the diff.

sparse_mask_training_image_only.png
Over successive training epochs, we prune the entries of matrices with too many non-zeroes, gradually sparsifying the diff.

When training the diff, we use an iterative procedure that prunes matrices with too many non-zero entries. As we did when training the arbitrator in the branched-encoder network, we turn up the temperature over successive epochs to make the diff sparser and sparser.

Our other approach to creating diffs was to use a hash function, a function that maps a large number of mathematical objects to a much smaller number of storage locations, or “buckets”. Hash functions are designed to distribute objects evenly across buckets, regardless of the objects’ values.

With this approach, we hash the locations in the diff matrix to buckets, and then, during training, we update the values in the buckets, rather than the values in the matrices. Since each bucket corresponds to multiple locations in the diff matrix, this reduces the amount of data we need to transfer to update a model. 

Hashed diffing.jpg
With hash diffing, a small number of weights (in the hash buckets at bottom) are used across a matrix with a larger number of entries.
Credit: Glynis Condon

One of the advantages of our approach, relative to other approaches to compression, such as matrix factorization, is that with each update, our diffs can target a different set of model weights. By contrast, traditional compression methods will typically lock you into modifying the same set of high-importance weights with each update.

AmazonScience_CarModel_V1.gif
An advantage of our diffing approach is that we can target a different set of weights with each model update, which gives us more flexibility in adapting to a changing data landscape.

In our experiments, we investigated the effects of three to five consecutive model updates, using different diffs for each. Hash diffing sometimes worked better for the first few updates, but over repeated iterations, models updated through hash diffing diverged more from full-parameter models. With sparsification diffing, the word error rate of a model updated five times in a row was less than 1% away from that of the full-parameter model, with diffs whose size was set at 10% of the full model’s.

Related content

US, WA, Seattle
Amazon.com strives to be Earth's most customer-centric company where customers can shop in our stores to find and discover anything they want to buy. We hire the world's brightest minds, offering them a fast paced, technologically sophisticated and friendly work environment. Economists at Amazon partner closely with senior management, business stakeholders, scientist and engineers, and economist leadership to solve key business problems ranging from Amazon Web Services, Kindle, Prime, inventory planning, international retail, third party merchants, search, pricing, labor and employment planning, effective benefits (health, retirement, etc.) and beyond. Amazon Economists build econometric models using our world class data systems and apply approaches from a variety of skillsets – applied macro/time series, applied micro, econometric theory, empirical IO, empirical health, labor, public economics and related fields are all highly valued skillsets at Amazon. You will work in a fast moving environment to solve business problems as a member of either a cross-functional team embedded within a business unit or a central science and economics organization. You will be expected to develop techniques that apply econometrics to large data sets, address quantitative problems, and contribute to the design of automated systems around the company. We are open to hiring candidates to work out of one of the following locations: Arlington, VA, USA | Bellevue, WA, USA | Boston, MA, USA | Los Angeles, CA, USA | New York, NY, USA | San Francisco, CA, USA | Seattle, WA, USA | Sunnyvale, CA, USA
US, WA, Seattle
Amazon.com strives to be Earth's most customer-centric company where customers can shop in our stores to find and discover anything they want to buy. We hire the world's brightest minds, offering them a fast paced, technologically sophisticated and friendly work environment. Economists at Amazon partner closely with senior management, business stakeholders, scientist and engineers, and economist leadership to solve key business problems ranging from Amazon Web Services, Kindle, Prime, inventory planning, international retail, third party merchants, search, pricing, labor and employment planning, effective benefits (health, retirement, etc.) and beyond. Amazon Economists build econometric models using our world class data systems and apply approaches from a variety of skillsets – applied macro/time series, applied micro, econometric theory, empirical IO, empirical health, labor, public economics and related fields are all highly valued skillsets at Amazon. You will work in a fast moving environment to solve business problems as a member of either a cross-functional team embedded within a business unit or a central science and economics organization. You will be expected to develop techniques that apply econometrics to large data sets, address quantitative problems, and contribute to the design of automated systems around the company. We are open to hiring candidates to work out of one of the following locations: Arlington, VA, USA | Bellevue, WA, USA | Boston, MA, USA | Los Angeles, CA, USA | New York, NY, USA | San Francisco, CA, USA | Seattle, WA, USA | Sunnyvale, CA, USA
US, WA, Seattle
Amazon.com strives to be Earth's most customer-centric company where customers can shop in our stores to find and discover anything they want to buy. We hire the world's brightest minds, offering them a fast paced, technologically sophisticated and friendly work environment. Economists at Amazon partner closely with senior management, business stakeholders, scientist and engineers, and economist leadership to solve key business problems ranging from Amazon Web Services, Kindle, Prime, inventory planning, international retail, third party merchants, search, pricing, labor and employment planning, effective benefits (health, retirement, etc.) and beyond. Amazon Economists build econometric models using our world class data systems and apply approaches from a variety of skillsets – applied macro/time series, applied micro, econometric theory, empirical IO, empirical health, labor, public economics and related fields are all highly valued skillsets at Amazon. You will work in a fast moving environment to solve business problems as a member of either a cross-functional team embedded within a business unit or a central science and economics organization. You will be expected to develop techniques that apply econometrics to large data sets, address quantitative problems, and contribute to the design of automated systems around the company. We are open to hiring candidates to work out of one of the following locations: Arlington, VA, USA | Bellevue, WA, USA | Boston, MA, USA | Los Angeles, CA, USA | New York, NY, USA | San Francisco, CA, USA | Seattle, WA, USA | Sunnyvale, CA, USA
LU, Luxembourg
The Decision, Science and Technology (DST) team part of the global Reliability Maintenance Engineering (RME) is looking for a Senior Operations Research Scientist interested in solving challenging optimization problems in the maintenance space. Our mission is to leverage the use of data, science, and technology to improve the efficiency of RME maintenance activities, reduce costs, increase safety and promote sustainability while creating frictionless customer experiences. As a Senior OR Scientist in DST you will be focused on leading the design and development of innovative approaches and solutions by leading technical work supporting RME’s Predictive Maintenance (PdM) and Spare Parts (SP) programs. You will connect with world leaders in your field and you will be tackling customer's natural language challenges by carrying out a systematic review of existing solutions. The appropriate choice of methods and their deployment into effective tools will be the key for the success in this role. The successful candidate will be a self-starter comfortable with ambiguity, with strong attention to detail and outstanding ability in balancing technical leadership with strong business judgment to make the right decisions about model and method choices. Key job responsibilities • Provide technical expertise to support team strategies that will take EU RME towards World Class predictive maintenance practices and processes, driving better equipment up-time and lower repair costs with optimized spare parts inventory and placement • Implement an advanced maintenance framework utilizing Machine Learning technologies to drive equipment performance leading to reduced unplanned downtime • Provide technical expertise to support the development of long-term spares management strategies that will ensure spares availability at an optimal level for local sites and reduce the cost of spares A day in the life As a Senior OR Scientist in DST you will be focused on leading the design and development of innovative approaches and solutions by leading technical work supporting RME’s Predictive Maintenance (PdM) and Spare Parts (SP) programs. You will connect with world leaders in your field and you will be tackling customer's natural language challenges by carrying out a systematic review of existing solutions. The appropriate choice of methods and their deployment into effective tools will be the key for the success in this role. About the team Our mission is to leverage the use of data, science, and technology to improve the efficiency of RME maintenance activities, reduce costs, increase safety and promote sustainability while creating frictionless customer experiences. We are open to hiring candidates to work out of one of the following locations: Luxembourg, LUX
US, WA, Seattle
Amazon.com strives to be Earth's most customer-centric company where customers can shop in our stores to find and discover anything they want to buy. We hire the world's brightest minds, offering them a fast paced, technologically sophisticated and friendly work environment. Economists at Amazon partner closely with senior management, business stakeholders, scientist and engineers, and economist leadership to solve key business problems ranging from Amazon Web Services, Kindle, Prime, inventory planning, international retail, third party merchants, search, pricing, labor and employment planning, effective benefits (health, retirement, etc.) and beyond. Amazon Economists build econometric models using our world class data systems and apply approaches from a variety of skillsets – applied macro/time series, applied micro, econometric theory, empirical IO, empirical health, labor, public economics and related fields are all highly valued skillsets at Amazon. You will work in a fast moving environment to solve business problems as a member of either a cross-functional team embedded within a business unit or a central science and economics organization. You will be expected to develop techniques that apply econometrics to large data sets, address quantitative problems, and contribute to the design of automated systems around the company. We are open to hiring candidates to work out of one of the following locations: Arlington, VA, USA | Bellevue, WA, USA | Boston, MA, USA | Los Angeles, CA, USA | New York, NY, USA | San Francisco, CA, USA | Seattle, WA, USA | Sunnyvale, CA, USA
US, WA, Seattle
Amazon.com strives to be Earth's most customer-centric company where customers can shop in our stores to find and discover anything they want to buy. We hire the world's brightest minds, offering them a fast paced, technologically sophisticated and friendly work environment. Economists in the Forecasting, Macroeconomics & Finance field document, interpret and forecast Amazon business dynamics. This track is well suited for economists adept at combining cutting edge times-series statistical methods with strong economic analysis and intuition. This track could be a good fit for candidates with research experience in: macroeconometrics and/or empirical macroeconomics; international macroeconomics; time-series econometrics; forecasting; financial econometrics and/or empirical finance; and the use of micro and panel data to improve and validate traditional aggregate models. Economists at Amazon are expected to work directly with our senior management and scientists from other fields on key business problems faced across Amazon, including retail, cloud computing, third party merchants, search, Kindle, streaming video, and operations. The Forecasting, Macroeconomics & Finance field utilizes methods at the frontier of economics to develop formal models to understand the past and the present, predict the future, and identify relevant risks and opportunities. For example, we analyze the internal and external drivers of growth and profitability and how these drivers interact with the customer experience in the short, medium and long-term. We build econometric models of dynamic systems, using our world class data tools, formalizing problems using rigorous science to solve business issues and further delight customers. We are open to hiring candidates to work out of one of the following locations: Arlington, VA, USA | Bellevue, WA, USA | Boston, MA, USA | Los Angeles, CA, USA | New York, NY, USA | San Francisco, CA, USA | Seattle, WA, USA | Sunnyvale, CA, USA
US, WA, Seattle
Amazon.com strives to be Earth's most customer-centric company where customers can shop in our stores to find and discover anything they want to buy. We hire the world's brightest minds, offering them a fast paced, technologically sophisticated and friendly work environment. Economists in the Forecasting, Macroeconomics & Finance field document, interpret and forecast Amazon business dynamics. This track is well suited for economists adept at combining cutting edge times-series statistical methods with strong economic analysis and intuition. This track could be a good fit for candidates with research experience in: macroeconometrics and/or empirical macroeconomics; international macroeconomics; time-series econometrics; forecasting; financial econometrics and/or empirical finance; and the use of micro and panel data to improve and validate traditional aggregate models. Economists at Amazon are expected to work directly with our senior management and scientists from other fields on key business problems faced across Amazon, including retail, cloud computing, third party merchants, search, Kindle, streaming video, and operations. The Forecasting, Macroeconomics & Finance field utilizes methods at the frontier of economics to develop formal models to understand the past and the present, predict the future, and identify relevant risks and opportunities. For example, we analyze the internal and external drivers of growth and profitability and how these drivers interact with the customer experience in the short, medium and long-term. We build econometric models of dynamic systems, using our world class data tools, formalizing problems using rigorous science to solve business issues and further delight customers. We are open to hiring candidates to work out of one of the following locations: Arlington, VA, USA | Bellevue, WA, USA | Boston, MA, USA | Los Angeles, CA, USA | New York, NY, USA | San Francisco, CA, USA | Seattle, WA, USA | Sunnyvale, CA, USA
US, WA, Seattle
Economists in the Forecasting, Macroeconomics & Finance field document, interpret and forecast Amazon business dynamics. This track is well suited for economists adept at combining cutting edge times-series statistical methods with strong economic analysis and intuition. This track could be a good fit for candidates with research experience in: macroeconometrics and/or empirical macroeconomics; international macroeconomics; time-series econometrics; forecasting; financial econometrics and/or empirical finance; and the use of micro and panel data to improve and validate traditional aggregate models. Economists at Amazon are expected to work directly with our senior management and scientists from other fields on key business problems faced across Amazon, including retail, cloud computing, third party merchants, search, Kindle, streaming video, and operations. The Forecasting, Macroeconomics & Finance field utilizes methods at the frontier of economics to develop formal models to understand the past and the present, predict the future, and identify relevant risks and opportunities. For example, we analyze the internal and external drivers of growth and profitability and how these drivers interact with the customer experience in the short, medium and long-term. We build econometric models of dynamic systems, using our world class data tools, formalizing problems using rigorous science to solve business issues and further delight customers. We are open to hiring candidates to work out of one of the following locations: Arlington, VA, USA | Bellevue, WA, USA | Boston, MA, USA | Los Angeles, CA, USA | New York, NY, USA | San Francisco, CA, USA | Seattle, WA, USA | Sunnyvale, CA, USA
US, WA, Seattle
Amazon.com strives to be Earth's most customer-centric company where customers can shop in our stores to find and discover anything they want to buy. We hire the world's brightest minds, offering them a fast paced, technologically sophisticated and friendly work environment. Economists at Amazon partner closely with senior management, business stakeholders, scientist and engineers, and economist leadership to solve key business problems ranging from Amazon Web Services, Kindle, Prime, inventory planning, international retail, third party merchants, search, pricing, labor and employment planning, effective benefits (health, retirement, etc.) and beyond. Amazon Economists build econometric models using our world class data systems and apply approaches from a variety of skillsets – applied macro/time series, applied micro, econometric theory, empirical IO, empirical health, labor, public economics and related fields are all highly valued skillsets at Amazon. You will work in a fast moving environment to solve business problems as a member of either a cross-functional team embedded within a business unit or a central science and economics organization. You will be expected to develop techniques that apply econometrics to large data sets, address quantitative problems, and contribute to the design of automated systems around the company. We are open to hiring candidates to work out of one of the following locations: Arlington, VA, USA | Bellevue, WA, USA | Boston, MA, USA | Los Angeles, CA, USA | New York, NY, USA | San Francisco, CA, USA | Seattle, WA, USA | Sunnyvale, CA, USA
US, WA, Seattle
Amazon.com strives to be Earth's most customer-centric company where customers can shop in our stores to find and discover anything they want to buy. We hire the world's brightest minds, offering them a fast paced, technologically sophisticated and friendly work environment. Economists at Amazon partner closely with senior management, business stakeholders, scientist and engineers, and economist leadership to solve key business problems ranging from Amazon Web Services, Kindle, Prime, inventory planning, international retail, third party merchants, search, pricing, labor and employment planning, effective benefits (health, retirement, etc.) and beyond. Amazon Economists build econometric models using our world class data systems and apply approaches from a variety of skillsets – applied macro/time series, applied micro, econometric theory, empirical IO, empirical health, labor, public economics and related fields are all highly valued skillsets at Amazon. You will work in a fast moving environment to solve business problems as a member of either a cross-functional team embedded within a business unit or a central science and economics organization. You will be expected to develop techniques that apply econometrics to large data sets, address quantitative problems, and contribute to the design of automated systems around the company. We are open to hiring candidates to work out of one of the following locations: Arlington, VA, USA | Bellevue, WA, USA | Boston, MA, USA | Los Angeles, CA, USA | New York, NY, USA | San Francisco, CA, USA | Seattle, WA, USA | Sunnyvale, CA, USA