Have you ever asked yourself how your smartphone recognizes what you say and who you are?Have you ever thought about how machines recognize different languages?
If you have an interesting idea on how to improve speech technology (for example, building an app that can truly transcribe an under-presented language, computing a model that can be easily tuned for African languages, or building a machine translation system for your under-represented community), then come and join us! The possibilities are endless! We need your creativity and enthusiasm to make it possible.
We are throwing a two-day speech and language technology hackathon during the IEEE Spoken Language Technology (SLT) Workshop in Doha, Qatar, on January 7th and 8th, 2023. It will be inspiring, momentous, and fun.
The goal of SLT-CODE is to build a diverse community of people who want to explore and envision how machines understand the spoken languages of the world. We believe this is best done by building something together! In the spirit of LANGUAGES OF THE WORLD we hope to encourage strong participation from regions that are currently underrepresented at language technology workshops and conferences to foster interest and connections across the world.
During the hackathon, you will focus on developing models, techniques, and applications for languages with limited resources (e.g., African, Asian, and Arabic languages).
This year's areas of interest for the hackathon are:
The organizers will give support to ASR, LID, MT and TTS.
It is open to new and seasoned speech researchers, programmers, user experience designers, practitioners, and speech and language technology enthusiasts.
In-person participation is preferred; however, remote participation is possible by joining a team with at least one person being local. Team sizes will be in the range of three to six.
The application process with the form below helps ensure we will have a good balance of participants given our limited capacity (around 100 people). We're looking for a broad swath of participants to create great projects relating to low-resource speech and language technology and applications. The hackathon is associated with the SLT 2022 workshop, but participation in the SLT Workshop is not required.
Apply by September 30, 2022 with this link: https://forms.gle/spXjyoXwwyo1jyYZ9
Now - September 30
Deadline to submit your incredible idea! Must contain:
October 12 - November 1:
November 1 to Jan 6:
Hackathon Jan 7-8:
We will be glad to meet you at:
What resources will be available:
There are many different frameworks that allow model creation for language technology.
Participants are free to use any technology platform they find suitable.
As organizers, we can offer support for the following frameworks:
The role of the mentor is to support teams with their perspective and experience. Each team will have a mentor assigned by November 1st. The mentor will be available for one hour every two weeks till the end of the Hackathon.
JUDGES AND JUDGING CRITERIA:
The judges will be experts in the field from academia and industry. The teams will be rated by the quality of the presentation (5 minutes) and optional demonstration (2 minutes), the novelty of the ideas explored, challenges that were overcome, and possible impact. To encourage the collaborative and community-building aspect of the Hackathon, teams will rank the top 5 most collaborative teams before the presentations.
All teams that give a presentation (and demonstration) will receive a certificate of participation. Teams can also receive an honorable mention which goes on their certificate. Further, some selected teams will be invited to give their presentation again at the SLT 2022 Workshop.
LOCATION AND DATES
Location: Doha, Qatar
Dates: Saturday, January 7th, 2023, and Sunday, January 8th, 2023
Recognition on: Monday Evening, January 9th, 2023.
We are in the process of looking for sponsors to help with travel costs for participants from underrepresented regions. If we get support this might take the form of shared hotel rooms, stipends, or support for travel expenses on an as-needed basis.