End of internship at VocaliD

Interning at VocaliD was definitely more than I expected it to be, and I was able to achieve my learning goals. The summer between my penultimate and final year was the perfect time for this opportunity, and I’ve come out of it with a greater sense of clarity when it comes to career paths I can pursue after graduation. A huge part of this was my career-specific goal of exposing myself to programming and its role in linguistics and speech science. For the past couple years at Brandeis I’ve considered more and more the option of pursuing further education in computational linguistics, and have become more interested in topics related to the field. The central role of speech science and text-to-speech technology in VocaliD’s work resonated with this interest, and has been all the convincing I need that this is a viable industry to attempt to enter in the coming years.

To another student looking for an internship at VocaliD, I would say this: be prepared for a fast-paced, interdisciplinary environment, and get ready to work with people of all calibers from all sorts of backgrounds. On more than one occasion there were company advisors in the office – often for advertising – and every one of them wanted to hear the opinion of the interns. Rather than sit back and simply absorb knowledge from experienced professionals, we were allowed to engage with them and be taken just as seriously.

This sort of open-mindedness could be an industry thing, or, more probably, due to the nature of small start-ups. There is a sense of urgency to everything that reinforces the “team” environment, requiring different, multi-faceted tasks from us on a daily basis. For this reason it felt very demanding, in a good way. The advice for somebody doing work for a tech start-up like this would be essentially the same, but phrased differently: the work you do is important, just as important as everyone else’s. This was by no means a “fetch coffee for the office” internship.

Emma, a fellow intern, and Sam, a previous recipient of a VocaliD voice, out for pizza in downtown Belmont.

Working for a company with a social mission was generally very rewarding. The effect we were having on people’s lives was so tangible, especially so when Samantha, a previous recipient of a VocaliD voice, came in to visit us at the office. Being able to see the difference in her regard for her old, generic voice and her new VocaliD voice put it in perspective how necessary the product is.

Maeve, a young girl with cerebral palsy, is receiving one of the voices we worked on this summer. Her story was featured heavily on our Indiegogo campaign.

And while my work this summer will go into voices that will be finished months from now, I am still proud to have participated in their creation. There are also customers awaiting their VocaliD voice currently (like Maeve, pictured above), and getting to see them receive it in the future is something I’m very excited for.

-David Stiefel ’16

Midpoint at VocaliD

Despite getting settled in a little more at VocaliD, my excitement about being here this summer hasn’t changed. It continues to be an interdisciplinary, dynamic environment, and though my central roles haven’t changed much, the details and everyday tasks vary from day to day, making it an always-interesting place to be. For a few weeks now, I have also been joined by another summer intern. The two of us work closely on some tasks and separately on others.

Our crowdfunding campaign ended its initial phase this week, marking a critical point for both the company and my summer. The campaign was largely a success, raising nearly twice as much as the initial goal, bringing in troves of new customers, and solidifying the coming timeline for VocaliD.

The current campaign status, showing our funding percentage.

For most of July, I’ll be doing more or less the same work as before, but applied towards the fulfillment of “perks” bought by customers on Indiegogo.

This week and next, we have a special visitor. Samantha Grimaldo was among the earliest recipients of a VocaliD voice, and an important pioneer in bringing the technology to market. We’re working with her to become more comfortable using her device to speak in public spaces, and documenting the process for a short video piece. Sam, who has much to say, made a few contributions to a new Tumblr page we’ve put up. In the future, the page will become an important place for people like Sam, who can feel very alone, to connect with other users of augmentative communication and share information about having and using a voice from VocaliD. In fact, most of the recipients of pre-orders during our campaign have been children or teenagers. The opportunity for somebody still young to be able to speak with their own voice is a wonderful thing to witness, and part of VocaliD’s service is that as your voice changes with age, the custom voice is updated to match these changes and always sound like you. Most of our past and current customers are young children, and that seems to be the demographic VocaliD most immediately affects. Hopefully the Tumblr page will allow for the sort of connecting between these young people that we hope.

In working on marketing-related things, such as drafting and sending email campaigns, I’ve become privy to just how many businesses today use email marketing as their main method of customer relations. We’ve been using Mailchimp, for example, and now I look at all of my email subscriptions and notice just how many companies use Mailchimp.

Mailchimp's ubiquitous email footer, common in emails we probably all subscribe to.

Email marketing is a staple today, especially for small businesses, and a great thing to have experience with, no matter what sort of business I may find myself in.

As someone who has worked a variety of jobs since early high school, I tend to think not of how work differs from university life, but more the other way around. In studying Linguistics and being exposed to academic publishing and field research, that always seems the more magical, less accessible, somewhat intimidating world that undergrads seem to mean when they talk about the “real world.” Even more daunting is the prospect of leading a life as a composer, which inherently involves connection with the academic world (and a good amount of financial struggle), and can be called “work” only in the loosest sense of the word.

Then again, that sort of thinking only reminds me that there aren’t really any such boundaries. My “world of work” this summer has been at a tech startup with a social mission, driven by donations from interested, generous people, and founded by a professor who underwent something of a STEM learning epiphany after some uplifting research findings. It blurs the lines between business and academia, something I often wish would happen to more of my university peers, and something I predict more of in the future business world.

-David Stiefel ’16

First Week at VocaliD

VocaliD, Inc. holds a very modern place in the business world. There is some amount of trouble capturing the operation in a succinct way, because paramount to VocaliD’s service to the augmentative and alternative communication (AAC) community is the data gathered from voice donors. The term “Socially-Oriented Company” has been getting thrown around more and more recently, and it is the most apt description of VocaliD’s nature, taking donated voices and using data from them to create ones for others in need.

The office is located on the third floor of the old firehouse in downtown Belmont, with a Pilates studio directly below and an Italian restaurant at street level. I love the location. There are plenty of places to grab good food for lunch, and the Fitchburg line station is a short walk away. On cooler mornings I bike in, which takes under a half hour.

I’ve been working alongside Rupal, the founder of the company, who is very easy to work with and a great supervisor. Most of my time this first week has been spent doing what I fully expected to be doing: examining, annotating, and editing speech data, in order to prepare it for the morphing algorithm VocaliD uses to create voices. However, we also launched a crowdfunding project on Indiegogo this week, and a lot of work went into designing and revising the campaign. I’ve also been writing portions of the various outreach emails that go out as part of the campaign and VocaliD’s business as usual. Going forward, tasks like these will continue to be part of my responsibilities this summer, so it looks like this internship will be getting me some interesting communications experience, from marketing to end users to forging relationships with other AAC companies.

If this week has been any indication of how the rest of the summer will be, then interning at VocaliD will be an incredible way of satisfying my WOW goals. I have the opportunity to work in a field that bridges signal processing and phonetics, two things I am familiar with from my two majors; I’m getting exposed to audio programming and code writing in a vocational setting, helping me to gain an understanding of programming and its place in computational linguistics; and VocaliD’s work presents a major, tangible service to those whose voices literally aren’t heard, and so I’m helping to eliminate inequalities faced daily by the AAC community.

Lastly, I’d like to talk about the logo and how well designed and appropriate it is, in addition to being tasteful and in line with current graphic design sensibilities.

the VocaliD company logo
The VocaliD company logo

At first glance, it’s a “V”, standing for all things vocal. Upon closer looking, the overall shape of the V is remarkably similar to that of human vocal folds. The graphic also consists visually of a small V inside a larger one, representing the way VocaliD blends just a few seconds of vocalization from a recipient along with several hours of donor speech to create the final product. The way in which these are overlaid, with alternating horizontal lines, is also very similar to the way waveforms of human vowels look, with secondary peaks and troughs layered inside.

waveform from a stereo recording of a young girl saying "thrown"
A waveform from a stereo recording of a young girl.

The logo has a whole lot of symbolism and information packed into it. It was partially designed by the founder herself, which is a great example of the interdisciplinary atmosphere of the whole team. This will, after all, be quite an interdisciplinary summer.

-David Stiefel ’16

Culminating my internship at CBRC

My summer at the Childhood Bilingualism Research Center was very fruitful; I accomplished all of my learning goals that I had set before beginning the internship. Time flew by as I worked on experiment design and data analysis, transcribed video files and learned how to use various programs and equipment at the Center. Additionally, I gained many new skills along the way that were beyond my expectations, such as learning how to use SPSS and the eye-tracker. Day by day, I became accustomed to the pace of working in academia alongside graduate students. All of these experiences will be useful for me in the future, academically and professionally.

I embarked on the internship with the academic goal of applying theoretical knowledge from my Brandeis courses to practical research. Originally only hoping to participate in experimental design, I actually got the chance to design an experiment from scratch. I created an interactive game studying trilingual children’s acquisition of spatial relations, making it fun for 4 to 6 year olds to participate in the study. In planning the experiment, I applied concepts from the language acquisition course I took this spring semester, and searched for relevant journal articles using databases introduced to me by a Brandeis professor. In my last week, I presented my ideas to the directors of the Center, Prof. Yip and Prof. Matthews, and all of the lab members.

Welcoming Dr. Gorter and Dr. Cenoz from University of the Basque Country, Spain

Throughout the internship, starting from the Conference in May, I met and chatted with many linguistics professors from around the world who came to visit the Center here in Hong Kong. It was eye-opening and refreshing to hear about the most recent studies about multilingual education and language policies across the globe. Like many others at the Center, I took pride that Hong Kong is becoming a vibrant academic meeting point where students and scholars come together to discuss the topic of multilingualism. I believe that this increased discourse will extend into the mainstream culture and encourage more parents to raise their children multilingually. Last year, CBRC collaborated with Radio Television Hong Kong to create a hour-long TV program promoting the positive outcomes of child multilingualism. This discussion has also been featured in an International Herald Tribune op-ed piece entitled “Cantonese, Please”.

Learning how to use the eye-tracker

For those who are interested in a research internship in linguistics, I really encourage you to connect with professors, in and out of Brandeis. Many of them are very keen to get to know undergraduates who are beginning a path in linguistics. Since there are so many sub-fields in linguistics, be sure to find a professor whose research interests align with your own. They may offer you an internship if you display passion for the subject and willingness to learn.

Dim sum with Prof. Yip, Prof. Matthews and Kenneth, visiting student from Harvard

During these past eight weeks, I gained valuable knowledge that will be important for me to have as I continue my studies in linguistics at Brandeis and explore possibilities for future research. I am so grateful to my supervisor and mentor Prof. Virginia Yip, without whom none of this would be possible and whose encouragement and guidance led me to challenge myself during the internship. Thank you to all of the lab members for introducing me to everything at the Center, showing me around the CUHK campus, and making sure I achieved all of my learning goals. Last but not least, thank you so much to the WOW committee for funding this very rewarding experience.

– Miriam Wong ’14

Moving from New England dialects to Hmong fieldwork

A lot has happened since my last blog entry. Besides working on more acoustic analysis, I made two trips down to Plymouth, New Hampshire to do some of my own interviews. I went with one of the Dartmouth students who I had met before.  He was very helpful in explaining exactly how he does the interviews, and we did the first one together. Then, I stayed at the bakery where we had set up, and he went off to other local spots where he thought he could get useful interviews. It was good for me to step out of my usual comfort zone and ask people who came in if they would be willing to be interviewed. I asked if they had grown up and lived most of their lives in the area, since that was what we were looking for.  If they answered “yes”, I told them a little about the project and asked if they had 8-10 minutes of time for an interview. I was lucky to receive mostly positive responses, and got about 10 interviews on my own within the two days. During the interview, I had them read a word list, reading passage, and sentences, followed by questions on whether they believed there is a New Hampshire or New England dialect. These interviews will be analyzed just like I have been analyzing previously conducted interviews, with Praat. An interesting thing I noticed when finding people to interview was that some people looked scary.  Yet,  I decided to approach them anyway, and they turned out to be the nicest ones. Among the various lessons I have learned, one is the typical, “don’t judge a book by its cover”! I have also refined my interviewing skills based upon this lesson.

The second day I went to interview people, I met a woman who had studied linguistics and who was very interested in the project. I gave her the Dartmouth professor’s business card, and she proceeded to contact him offering to help with the project, which he was very excited about! He appreciated my personable attitude and said that he believed I would do great on the Hmong project, as it seemed like I was very approachable. I felt proud that I could be such a help to the project, and the interviews made me feel as if I was a valuable component; more so than when I was simply doing analysis from home.

During the remainder of my time at home, the professor also gave me books to look through about the Hmong. I had previously read Anne Fadiman’s book, “The Spirit Catches You and You Fall Down,” but besides that, did not know much about the community. I have already learned a lot more about them simply through the books. A lot of the material talked about the fact that many Hmong who now live in America feel as if Americans do not understand their culture, and misinterpret many cultural rituals and traditions. One thing I have noticed when reading these books is that it is much easier for me to retain the information when I am studying it for my own use, rather than simply for a test in class. I am excited to learn even more as I actually begin to interview the Hmong people.

“The Mong Oral Tradition” – A few of the books that the Dartmouth professor has provided me with.

 

I just got to Massachusetts yesterday, where I will be spending the remainder of my internship. Before I left, I stopped at Dartmouth to speak with the professor about what exactly I will be doing during my time here, since the work is mostly on my own. He suggested I contact the Brandeis student again who did Hmong field work a few years ago. He also gave me all of her previous Hmong contacts, notes and interviews. I have already contacted her and she told me which places she was most successful, most of which were in Providence, RI, though also one park where she met a lot of Hmong people in Fitchburg, MA. Otherwise, I should begin by researching online to find Hmong organizations in the area, as it very well may have changed a bit since the previous Brandeis student carried out fieldwork here. Once I start conducting interviews, they will include cultural questions as well as certain components that will allow the interviewees to speak Hmong, which we can analyze later to find interesting linguistic elements within the language.

I am nervous because I feel even more on my own now than before, but the professor is more than helpful in answering any questions, and I feel as if I am well prepared. He will check in with me every week to make sure I am doing well with the research, and he will either visit me here at some point, or I will make a trip back to speak with him and possibly even do some more of the New England dialect field work. And whenever I am not busy with Hmong work, there is always more acoustic analysis to be done! The professor has assured me that even if I do not make a life-changing discovery, making more Hmong contacts in the area and carrying out some interviews will be very helpful to him. And personally, I have already learned so much that I know this internship has been and will continue to be beneficial to me! I am learning skills both that I can use in life, and more specific skills that I can use for future linguistics work. Although I am about half way through, I am only beginning this part of the internship, and even though I am nervous I am also so excited to see what will happen!

Me working in my new room! Trying to beat the heat…

– Alexandra Patch ’14

Researching Cantonese-English-Mandarin Language Acquisition

More than halfway into my internship, I have been making good progress on my learning goals for the summer at the Childhood Bilingualism Research Centre. In order to monitor my growth and make sure that I am absorbing as much knowledge as possible here, I have been keeping track of my completed tasks and constantly asking the graduate students at the Center for feedback and comments.

At the moment, I am transcribing video recordings for the Hong Kong Bilingual Child Language Corpus. The corpus is a database of bilingual Cantonese-English child speech recordings, in audio and video formats. I use the linguistic software called CLAN to transcribe and mark down specific features that appear in the child’s speech. For example, an important feature to note is code-switching, which is when the child switches from language to another, in this case from Cantonese to English or vice versa. Moreover, we not only transcribe the speech production of the child subject, but just as importantly, the production of the adults who speak to the child, or in other words, the child’s language input. We can achieve a more complete understanding of the target child’s language attainment by examining both her linguistic input and output.

Transcribing target child Yarona’s (mainly) English conversation — click to expand!

Another project I have been working on is the Mandarin Receptive Vocabulary Test for Hong Kong Children. One of my goals for the summer was to conduct experiments that look at children’s acquisition of vocabulary and sentences to better understand how teachers and curriculum can provide more effective language education for children. My responsibility was to compile the results and calculated the scores of each child who took the MRVT. The test is given to children aged 4-6 to assess their acquisition of Mandarin, a second or third language after Cantonese and English for most children in Hong Kong. In the test, children hear a word spoken in Mandarin and are asked to point to the corresponding picture. Only one out of the four pictures is correct and the other options are carefully selected distractors. There is always one other picture that is similar sounding, and one that is similar in meaning. The results tell us how children are most likely to make mistakes, and indicate areas that parents and teachers can improve upon. Working on this project gave me a lot of insight into my long-term goal which is to pursue a career incorporating linguistics into education, so that children can be exposed to various languages at an early age to become global citizens when they grow up. They will be able to communicate with many people, yet also have a native language that reminds them of their heritage.

Sample question in the MRVT: which picture shows xiang1 jiao1?

Concurrently with the other projects, I am currently working to design a computer-based experiment to study the referential strategy of spatial relations. It is extremely challenging and I get a great deal of independence in researching and designing how the experiment will be set up and run. It requires a lot of creative thinking and research. I am learning about the scientific method and research process. Working at CBRC, I have gained skills that will be essential for me in the future. Specifically, I have gained skills in transcription, and am working at a much faster pace than when I first started.

– Miriam Wong ’14

My first week at Dartmouth!

It’s been a little over a week since I began my internship, but there’s been so much going on that I only have time now to sit down and write this blog. So far, my internship has been great, and is definitely meeting my expectations. The first day, I actually had to do an online training called CITI, or the “Collaborative Institutional Training Initiative,” because I’ll be working with human subjects. It took multiple hours, but now I’m proud to say I’m CITI certified! The next day, I went to Dartmouth to meet my supervisor, a professor of Linguistics. The plan was for me to work on two of his projects;  carrying out field work in New England, as well as working from home or at the phonetics lab on acoustic analysis.  We had corresponded previously by email and phone, and it was very exciting to meet him. Since then, I’ve gone two more times, one to see the seniors’ linguistic thesis presentations, and once because my responsibilities include checking up with my supervisor once a week. At the thesis presentations, I met linguistic students at Dartmouth with whom I will be working on the New England dialect project. This project includes traveling around New Hampshire and Vermont and interviewing people in order to listen to their dialects. The students were really nice, and I’m excited to travel with them! I’m glad that I will be able to interact with other people my age, because at first I thought that it might all be on my own.

Dartmouth during my first visit!

In the phonetics lab I have started the acoustic analysis of people’s dialects from both Vermont and New Hampshire from previous fieldwork recordings.  I use software called “Praat” to analyze speech. Right now I’m focusing specifically on vowels and whether or not people pronounce “r’s” in words (this is called “Rhoticity”). I record the data in Excel, and use another program called “StatPlus” to analyze it further.  I have already learned so many valuable linguistic skills, and I am excited to learn even more! This screenshot is an example of the work I’ve been doing. The red dots are called “formants,” and I record the Hz of the two bottom ones, which become F1 and F2. Charted, this can be compared to standard English, and can determine whether a person’s dialect is different from standard English.

Screenshot of "Praat"...analyzing the vowel in "law"

Later on in the summer I will move to the Boston area to carry out fieldwork with Hmong, an Asian ethnic group, many of whom have immigrated to the US. This will be exciting for me, as I have read about them in my courses at Brandeis. My supervisor has given me books to read about their culture, and I’m looking forward to learning more through these accounts.  I found this internship through the “Brandeis Internship Exchange,” as someone had done Hmong work with the same professor three years ago. This sparked my interest, and I decided to contact the professor to see if he had any need of an intern at this point. He was very excited to hear from me, and after corresponding about my preparedness through coursework at Brandeis, and his available projects, we decided on the two projects that seemed to fit me best.

At the beginning of my internship I was worried that I would be working alone. However, my supervisor is very helpful discussing expectations during our weekly meetings. I set daily goals for myself and I am able to do the amount of work he expects me to do. I am excited to continue this internship, and I really can’t believe how much I’ve learned already. So far, I’d actually say it has exceeded my expectations. I feel like I can only learn more from this point on, and this is really showing me that linguistics is a field I would like to pursue!

– Alexandra Patch ’14