Amazon Text-to-Speech group's research at ICASSP 2022

Papers focus on speech conversion and data augmentation — and sometimes both at once.

The automatic conversion of text to speech is crucial to Alexa: it’s how Alexa communicates with customers. The models developed by the Amazon Text-to-Speech group are also available to Amazon Web Services (AWS) customers through Polly, the AWS text-to-speech service.

The Text-to-Speech (TTS) group has four papers at this year’s International Conference on Acoustics, Speech, and Signal Processing (ICASSP), all of which deal with either voice conversion (preserving prosodic features while converting one synthetic voice to another), data augmentation, or both.

More ICASSP coverage on Amazon Science

In “Voice Filter: Few-shot text-to-speech speaker adaptation using voice conversion as a post-processing module”, the Amazon TTS group addresses the problem of few-shot speaker adaptation, or learning a new synthetic voice from just a handful of training examples. The paper reformulates the problem as learning a voice conversion model that’s applied to the output of a high-quality TTS model, a conceptual shift from the existing few-shot-TTS paradigm.

In “Cross-speaker style transfer for text-to-speech using data augmentation”, the team shows how to build a TTS model capable of expressive speech, even when the only available training data for the target voice consists of neutral speech. The idea is to first train a voice conversion model, which converts samples of expressive speech in other voices into the target voice, and then use the converted speech as additional training data for the TTS model.

In “Distribution augmentation for low-resource expressive text-to-speech”, the TTS group expands the range of texts used to train a TTS model by recombining excerpts from existing examples to produce new examples. The trick is to maintain the syntactic coherence of the synthetic examples, so that the TTS model won’t waste resources learning improbable sequences of phonemes. (This is the one data augmentation paper that doesn’t rely on voice conversion.)

Parse substitution.png
In this example of data augmentation through recombination of existing training examples, the verb phrase “shook her head”, as identified by a syntactic parse, is substituted for the verb phrase “lied” in the sentence “he never lied”. The original acoustic signals (bottom row) are cut and spliced at the corresponding points. From "Distribution augmentation for low-resource expressive text-to-speech".

Finally, in “Text-free non-parallel many-to-many voice conversion using normalising flows”, the team adapts the concept of normalizing flows, which have been used widely for TTS, to the problem of voice conversion. Like most deep-learning models, normalizing flows learn functions that produce vector representations of input data. The difference is that the functions are invertible, so the inputs can be recovered from the representations. The team hypothesized that preserving more information from the input data would yield better voice conversion, and early experiments bear that hypothesis out.

Voice filter

The idea behind “Voice Filter: Few-shot text-to-speech speaker adaptation using voice conversion as a post-processing module” is that for few-shot learning, it’s easier to take the output of an existing, high-quality TTS model — a voice spectrogram — and adapt that to a new target voice than it is to adapt the model itself.

The key to the approach is that the voice filter, which converts the TTS model’s output to a new voice, is trained on synthetic data created by the TTS model itself.

Voice filter.png
The training procedure for the voice filter.

The TTS model is duration controllable, meaning that the input text is encoded to indicate the duration that each phoneme should have in the output speech. This enables the researchers to create two parallel corpora of training data. One corpus consists of real training examples, from 120 different speakers. The other corpus is synthetic speech generated by the TTS model, but with durations that match those of the multispeaker examples.

The voice filter is trained on the parallel corpora, and then, for few-shot learning, the researchers simply fine-tune it on a new speaker. In experiments, the researchers found that this approach produced speech whose quality was comparable to that produced by conventional models trained on 30 times as much data.

Cross-speaker style transfer

The voice conversion model that the researchers use in “Cross-speaker style transfer for text-to-speech using data augmentation” is based on the CopyCat model previously reported on the Amazon Science blog. The converted expressive data is added to the neutral data to produce the dataset used to train the TTS model

The TTS model takes two inputs: a text sequence and a style vector. During training, the text sequence passes to the TTS model, and the spectrogram of the target speech sample passes to a reference encoder, which produces the style embedding. At inference time, of course, there is no input spectrogram. But the researchers show that they can control the style of the TTS model’s output by feeding it a precomputed style embedding.

Cross-speaker style transfer.png
The voice conversion model (left) and text-to-speech model (right) used for cross-speaker style transfer. The reference encoders are used only during training. From "Cross-speaker style transfer for text-to-speech using data augmentation".

The researchers assessed the model based on human evaluation using the MUSHRA perception scale. Human evaluators reported that, relative to a benchmark model, the new model reduced the gap in perceived style similarity between synthesized and real speech by an average of 58% across 14 different speakers.

Distribution augmentation

Distribution augmentation for low-resource expressive text-to-speech” considers the case in which training data for a new voice is lacking. The goal is to permute the texts of the existing examples, producing new examples, and recombine excerpts from the corresponding speech samples to produce new samples. This does not increase the acoustic diversity of the training targets, but it does increase the linguistic diversity of the training inputs.

To ensure that the synthetic training examples do not become too syntactically incoherent, the researchers construct parse trees for the input texts and then swap syntactically equivalent branches across trees (see figure, above). Swapping the corresponding sections of the acoustic signal requires good alignment between text and signal, which is accomplished by existing forced-alignment models.

During training, to ensure that the resulting TTS model doesn’t become overbiased toward the synthetic examples, the researchers also include a special input token to indicate points at which two existing samples have been fused together. The expectation is that the model will learn to privilege phonemic sequences internal to the real samples over phonemic sequences that cross boundaries between fused samples. At inference time, the value of the token is simply set to 0 across all inputs.

Fused-sample token.png
The “augmentation tag” marks the boundary between acoustic signals taken from two different training examples, to prevent overbiasing the TTS model toward synthetic data. From "Distribution augmentation for low-resource expressive text-to-speech".

The quality of the model’s speech output was assessed by 60 human evaluators, who compared it to speech output by a baseline model, on five different datasets. Across the board, the output of the new model received better scores than the output of the benchmark model.

Normalizing flows

A normalizing flow learns to map input data to a representational space in a way that maximizes the approximation of some prior distribution. The word “flow” indicates that the mapping can be the result of passing the data through a series of invertible transformations, and the enforcement of the distribution imposes the normalization.

In “Text-free non-parallel many-to-many voice conversion using normalising flows”, Amazon TTS researchers consider a flow whose inputs are a source spectrogram, a phoneme embedding, a speaker identity embedding, the fundamental frequency of the acoustic signal, and a flag denoting whether a frame of input audio is voiced or unvoiced. The flow maps the inputs to a distribution of phoneme frequencies in a particular application domain.

Typically, a normalizing flow will learn both the distribution and the mapping from the training data. But here, the researchers pretrain the flow on a standard TTS task, for which training data is plentiful, to learn the distribution in advance.

Because the flow is reversible, a vector in the representational space can be mapped back to a set of source inputs, provided that the other model inputs (phoneme embedding, speaker ID, and so on) are available. To use normalizing flows to perform speech conversion, the researchers simply substitute one speaker for another during this reverse mapping.

Normalizing flow.png
An overview of TTS researchers' use of normalizing flows to do voice conversion. From "Text-free non-parallel many-to-many voice conversion using normalising flows".

The researchers examine two different experimental setting, one in which the voice conversion model takes both text sequences and spectrograms as inputs and one in which it takes spectrograms only. In the second case, the pretrained normalizing-flow model significantly outperformed the benchmarks. A normalizing-flow model that learned the phoneme distribution directly from the training data didn’t fare as well, indicating the importance of the pretraining step.

Related content

US, VA, Herndon
Do you love decomposing problems to develop machine learning (ML) products that impact millions of people around the world? Would you enjoy identifying, defining, and building ML software solutions that revolutionize how businesses operate? The Global Practice Organization in Professional Services at Amazon Web Services (AWS) is looking for a Software Development Engineer II to build, deliver, and maintain complex ML products that delight our customers and raise our performance bar. You’ll design fault-tolerant systems that run at massive scale as we continue to innovate best-in-class services and applications in the AWS Cloud. Key job responsibilities Our ML Engineers collaborate across diverse teams, projects, and environments to have a firsthand impact on our global customer base. You’ll bring a passion for the intersection of software development with generative AI and machine learning. You’ll also: - Solve complex technical problems, often ones not solved before, at every layer of the stack. - Design, implement, test, deploy and maintain innovative ML solutions to transform service performance, durability, cost, and security. - Build high-quality, highly available, always-on products. - Research implementations that deliver the best possible experiences for customers. A day in the life As you design and code solutions to help our team drive efficiencies in ML architecture, you’ll create metrics, implement automation and other improvements, and resolve the root cause of software defects. You’ll also: - Build high-impact ML solutions to deliver to our large customer base. - Participate in design discussions, code review, and communicate with internal and external stakeholders. - Work cross-functionally to help drive business solutions with your technical input. - Work in a startup-like development environment, where you’re always working on the most important stuff. About the team The Global Practice Organization for Analytics is a team inside the AWS Professional Services Organization. Our mission in the Global Practice Organization is to be at the forefront of defining machine learning domain strategy, and ensuring the scale of Professional Services' delivery. We define strategic initiatives, provide domain expertise, and oversee the development of high-quality, repeatable offerings that accelerate customer outcomes. Inclusive Team Culture Here at AWS, we embrace our differences. We are committed to furthering our culture of inclusion. We have thirteen employee-led affinity groups, reaching 85,000 employees in over 190 chapters globally. We have innovative benefit offerings, and host annual and ongoing learning experiences, including our Conversations on Race and Ethnicity (CORE) and AmazeCon (gender diversity) conferences. Amazon’s culture of inclusion is reinforced within our 16 Leadership Principles, which remind team members to seek diverse perspectives, learn and be curious, and earn trust. Work/Life Balance Our team puts a high value on work-life harmony. Striking a healthy balance between your personal and professional life is crucial to your happiness and success here. We are a customer-obsessed organization—leaders start with the customer and work backwards. They work vigorously to earn and keep customer trust. As such, this is a customer facing role in a hybrid delivery model. Project engagements include remote delivery methods and onsite engagement that will include travel to customer locations as needed. Mentorship & Career Growth Our team is dedicated to supporting new members. We have a broad mix of experience levels and tenures, and we’re building an environment that celebrates knowledge sharing and mentorship. We care about your career growth and strive to assign projects based on what will help each team member develop into a better-rounded professional and enable them to take on more complex tasks in the future. This is a customer-facing role and you will be required to travel to client locations and deliver professional services as needed. We are open to hiring candidates to work out of one of the following locations: Atlanta, GA, USA | Austin, TX, USA | Boston, MA, USA | Chicago, IL, USA | Herndon, VA, USA | Minneapolis, MN, USA | New York, NC, USA | San Diego, CA, USA | San Francisco, CA, USA | Seattle, WA, USA
US, MA, North Reading
Are you inspired by invention? Is problem solving through teamwork in your DNA? Do you like the idea of seeing how your work impacts the bigger picture? Answer yes to any of these and you’ll fit right in here at Amazon Robotics. We are a smart team of doers that work passionately to apply cutting edge advances in robotics and software to solve real-world challenges that will transform our customers’ experiences in ways we can’t even imagine yet. We invent new improvements every day. We are Amazon Robotics and we will give you the tools and support you need to invent with us in ways that are rewarding, fulfilling and fun. Amazon Robotics is seeking Applied Science Interns and Co-ops with a passion for robotic research to work on cutting edge algorithms for robotics. Our team works on challenging and high-impact projects within robotics. Examples of projects include allocating resources to complete a million orders a day, coordinating the motion of thousands of robots, autonomous navigation in warehouses, identifying objects and damage, and learning how to grasp all the products Amazon sells. As an Applied Science Intern/Co-op at Amazon Robotics, you will be working on one or more of our robotic technologies such as autonomous mobile robots, robot manipulators, and computer vision identification technologies. The intern/co-op project(s) and the internship/co-op location are determined by the team the student will be working on. Please note that by applying to this role you would be considered for Applied Scientist summer intern, spring co-op, and fall co-op roles on various Amazon Robotics teams. These teams work on robotics research within areas such as computer vision, machine learning, robotic manipulation, navigation, path planning, perception, optimization and more. Learn more about Amazon Robotics: https://amazon.jobs/en/teams/amazon-robotics We are open to hiring candidates to work out of one of the following locations: North Reading, MA, USA | Seattle, WA, USA | Westborough, MA, USA
CA, BC, Vancouver
Amazon Web Services (AWS) is building a world-class marketing organization that drives awareness and customer engagement with the goal of educating developers, IT and line-of-business professionals, startups, partners, and executive decision makers about AWS services and solutions, their benefits, and differentiation. As the central data and science organization in AWS Marketing, the Data: Science and Engineering (D:SE) team builds measurement products, AI/ML models for targeting, and self-service insights capabilities for AWS Marketing to drive better measurement and personalization, improve data access and analytical self-service, and empower strategic data-driven decisions. We work globally as a central team and establish standards, benchmarks, and best practices for use throughout AWS Marketing. We are looking for a Principal Data Scientist with deep expertise in scaling measurement science, content ranking and rapid experimentation at scale, with strong interest in building scalable solutions in partnership with our engineering organization. You will lead strategic measurement science initiatives across AWS Marketing & Sales ranging anywhere between recommender engines, scaling experimentation and measurement science, real-time inference, and cross-channel orchestration. You are an hands-on innovator who can contribute to advancing Marketing measurement technology in a B2B environment, and push the limits on what’s scientifically possible with a razor sharp focus on measurable customer and business impact. You will work with recognized B2B Marketing Science and AI/ML experts to develop large-scale, high-performing measurement science models and AI/ML capabilities. We are at a pivotal moment in our organization where AI/ML and measurement velocity has reached an unseen momentum, and we need to scale fast in order to maintain it. Your work will be a key input into a few of our key business goals. You will advance the state of the art in measurement at scale. We are open to hiring candidates to work out of one of the following locations: Vancouver, BC, CAN
US, WA, Seattle
Innovators wanted! Are you an entrepreneur? A builder? A dreamer? This role is part of an Amazon Special Projects team that takes the company’s Think Big leadership principle to the extreme. We focus on creating entirely new products and services with a goal of positively impacting the lives of our customers. No industries or subject areas are out of bounds. If you’re interested in innovating at scale to address big challenges in the world, this is the team for you. Here at Amazon, we embrace our differences. We are committed to furthering our culture of inclusion. We have thirteen employee-led affinity groups, reaching 40,000 employees in over 190 chapters globally. We are constantly learning through programs that are local, regional, and global. Amazon’s culture of inclusion is reinforced within our 16 Leadership Principles, which remind team members to seek diverse perspectives, learn and be curious, and earn trust. Our team highly values work-life balance, mentorship and career growth. We believe striking the right balance between your personal and professional life is critical to life-long happiness and fulfillment. We care about your career growth and strive to assign projects and offer training that will challenge you to become your best. Key job responsibilities • Develop automated laboratory workflows. • Perform data QC, document results, and communicate to stakeholders. • Maintain updated understanding and knowledge of methods. • Identify and escalate equipment malfunctions; troubleshoot common errors. • Participate in the updating of protocols and database to accurately reflect the current practices. • Maintain equipment and instruments in good operating condition • Adapt to unexpected schedule changes and respond to emergency situations, as needed. We are open to hiring candidates to work out of one of the following locations: Seattle, WA, USA
US, WA, Seattle
Are you excited about developing generative AI and foundation models to revolutionize automation, robotics and computer vision? Are you looking for opportunities to build and deploy them on real problems at truly vast scale? At Amazon Fulfillment Technologies and Robotics we are on a mission to build high-performance autonomous systems that perceive and act to further improve our world-class customer experience - at Amazon scale. We are looking for scientists, engineers and program managers for a variety of roles. The Amazon Robotics software team is seeking a Applied Scientist to focus on large vision and manipulation machine learning models. This includes building multi-viewpoint and time-series computer vision systems. It includes using machine learning to drive hardware movement. It includes building large-scale models using data from many different tasks and scenes. This work spans from basic research such as cross domain training, to experimenting on prototype in the lab, to running wide-scale A/B tests on robots in our facilities. Key job responsibilities * Research vision - Where should we be focusing our efforts * Research delivery – Proving/dis-proving strategies in offline data or in the lab * Production studies - Insights from production data or ad-hoc experimentation. About the team This team invents and runs robots focused on grasping and packing items. These are typically 6-dof style robotic arms. Our work ranges from the long-term-research on basic science to deploying/supporting large production fleets handling billions of items per year. We are open to hiring candidates to work out of one of the following locations: Seattle, WA, USA
US, VA, Arlington
Amazon launched the Generative AI (GenAI) Innovation Center (GAIIC) in Jun 2023 to help AWS customers accelerate enterprise innovation and success with Generative AI (https://press.aboutamazon.com/2023/6/aws-announces-generative-ai-innovation-center). Customers such as Highspot, Lonely Planet, Ryanair, and Twilio are engaging with the GAI Innovation Center to explore developing generative solutions. GAIIC provides opportunities to innovate in a fast-paced organization that contributes to game-changing projects and technologies that get deployed on devices and in the cloud. As a data scientist at GAIIC, you are proficient in designing and developing advanced Generative AI based solutions to solve diverse customer problems. You will be working with terabytes of text, images, and other types of data to solve real-world problems through Gen AI. You will be working closely with account teams and ML strategists to define the use case, and with other scientists and ML engineers on the team to design experiments, and find new ways to deliver value to the customer. The successful candidate will possess both technical and customer-facing skills that will allow you to be the technical “face” of AWS within our solution providers’ ecosystem/environment as well as directly to end customers. You will be able to drive discussions with senior technical and management personnel within customers and partners. This position requires that the candidate selected be a US Citizen and currently possess and maintain an active Top Secret security clearance. About the team Work/Life Balance Our team puts a high value on work-life balance. It isn’t about how many hours you spend at home or at work; it’s about the flow you establish that brings energy to both parts of your life. We believe striking the right balance between your personal and professional life is critical to life-long happiness and fulfillment. We offer flexibility in working hours and encourage you to find your own balance between your work and personal lives. Mentorship & Career Growth Our team is dedicated to supporting new members. We have a broad mix of experience levels and tenures, and we’re building an environment that celebrates knowledge sharing and mentorship. Our senior members enjoy one-on-one mentoring and thorough, but kind, code reviews. We care about your career growth and strive to assign projects based on what will help each team member develop into a better-rounded engineer and enable them to take on more complex tasks in the future. We are open to hiring candidates to work out of one of the following locations: Arlington, VA, USA | Denver, CO, USA
US, VA, Arlington
Amazon’s mission is to be the most customer centric company in the world. The Workforce Staffing (WFS) organization is on the front line of that mission by hiring the hourly fulfillment associates who make that mission a reality. To drive the necessary growth and continued scale of Amazon’s associate needs within a constrained employment environment, Amazon has created the Workforce Intelligence (WFI) team. This team will (re)invent how Amazon attracts, communicates with, and ultimately hires its hourly associates. This team owns multi-layered research and program implementation to drive deep learning, process improvements, and strategic recommendations to global leadership. Are you passionate about data? Do you enjoy questioning the status quo? Do complex and difficult challenges excite you? If yes, this may be the team for you. The Data Scientist will be responsible for creating cutting edge algorithms, predictive and prescriptive models as well as required data models to facilitate WFS at-scale warehouse associate hiring. This role acts as an internal consultant to the marketing, biz ops and candidate experience teams covering responsibilities such as at-scale hiring process improvement, analyzing large scale candidate/associate data and being strategic to providing best candidate hiring experience to WFS warehouse associate candidates. We are open to hiring candidates to work out of one of the following locations: Arlington, VA, USA
US, WA, Seattle
Innovators wanted! Are you an entrepreneur? A builder? A dreamer? This role is part of an Amazon Special Projects team that takes the company’s Think Big leadership principle to the extreme. We focus on creating entirely new products and services with a goal of positively impacting the lives of our customers. No industries or subject areas are out of bounds. If you’re interested in innovating at scale to address big challenges in the world, this is the team for you. Here at Amazon, we embrace our differences. We are committed to furthering our culture of inclusion. We have thirteen employee-led affinity groups, reaching 40,000 employees in over 190 chapters globally. We are constantly learning through programs that are local, regional, and global. Amazon’s culture of inclusion is reinforced within our 16 Leadership Principles, which remind team members to seek diverse perspectives, learn and be curious, and earn trust. Our team highly values work-life balance, mentorship and career growth. We believe striking the right balance between your personal and professional life is critical to life-long happiness and fulfillment. We care about your career growth and strive to assign projects and offer training that will challenge you to become your best. We are open to hiring candidates to work out of one of the following locations: Seattle, WA, USA
US, CA, Sunnyvale
Are you passionate about solving unique customer-facing problem at Amazon scale? Are you excited by developing and productionizing machine learning, deep learning algorithms and leveraging tons of Amazon data to learn and infer customer shopping patterns? Do you enjoy working with a diverse set of engineers, machine learning scientists, product managers and user-experience designers? If so, you have found the right match! Virtual Try On (VTO) at Amazon Fashion & Fitness is looking for an exceptional Applied Scientist to join us to build our next generation virtual try on experience. Our goal is to help customers evaluate how products will fit and flatter their unique self before they ship, transforming customers' shopping into a personalized journey of inspiration, discovery, and evaluation. In this role, you will be responsible for building scalable computer vision and machine learning (CVML) models, and automating their application and expansion to power customer-facing features. Key job responsibilities - Tackle ambiguous problems in Computer Vision and Machine Learning, and drive full life-cycle of CV/ML projects. - Build Computer Vision, Machine Learning and Generative AI models, perform proof-of-concept, experiment, optimize, and deploy your models into production. - Investigate and solve exciting and difficult challenges in Image Generation, 3D Computer Vision, Generative AI, Image Understanding and Deep Learning. - Run A/B experiments, gather data, and perform statistical tests. - Lead development and productionalization of CV, ML, and Gen AI models and algorithms by working across teams. Deliver end to end. - Act as a mentor to other scientists on the team. We are open to hiring candidates to work out of one of the following locations: Sunnyvale, CA, USA
US, CA, Sunnyvale
At Amazon Fashion, we are obsessed with making Amazon Fashion the most loved fashion destinations globally. We're searching for Computer Vision pioneers who are passionate about technology, innovation, and customer experience, and who are enthusiastic about making a lasting impact on the industry. You'll be working with talented scientists, engineers, and product managers to innovate on behalf of our customers. If you're fired up about being part of a dynamic, driven team, then this is your moment to join us on this exciting journey and change the world of eCommerce forever Key job responsibilities As a Applied Scientist, you will be at the forefront to define, own and drive the science that span multiple machine learning models and enabling multiple product/engineering teams and organizations. You will partner with product management and technical leadership to identify opportunities to innovate customer facing experiences. You will identify new areas of investment and work to align product roadmaps to deliver on these opportunities. As a science leader, you will not only develop unique scientific solutions, but more importantly influence strategy and outcomes across different Amazon organizations such as Search, Personalization and more. This role is inherently cross-functional and requires a strong ability to communicate, influence and earn the trust of software engineers, technical and business leadership. We are open to hiring candidates to work out of one of the following locations: Sunnyvale, CA, USA