Research @awscloud ፨ Graduated @FBK_MT @UniTrento ፨ 💚💛❤️
San Francisco Bay Area
Joined November 2009
- Tweets1.2K
- Following371
- Followers410
- Likes5.2K
Surafel | ሱራፌል retweeted
Proceedings for IWSLT shared task are now online aclanthology.org/events/iwsl…
Our #ICASSP paper, Isometric MT: Neural Machine Translation for Automatic Dubbing is out.
amazon.science/publications/…
We show how self-learning is crucial to control the output of translation models.
Discover how we applied the proposed approach in an automatic dubbing use case. #AmazonScience
Working on #Neural #Machine #Translation
Consider participating in this year @iwslt Isometric translation task. More details 👇
We are organizing Isometric MT shared task in @iwslt 2022 iwslt.org/2022/isometric
What's the problem?
Generate translations which are similar in length to source.
(1/3)
Surafel | ሱራፌል retweeted
The ✨ 2nd AfricaNLP Workshop ✨ will be taking place at #EACL2021! They have a magnificent programme 🤩
Please do submit 🌠 So we can build together 🌍
📅 Submission date: 5 Feb 2021
📝 Short (4pg) or long (8pg) papers
👇🏿All details within👇🏿
sites.google.com/view/africa…
Surafel | ሱራፌል retweeted
The 3rd Workshop on technologies for Low Resource languages #LoResMT starts in a few hours.
Check out the rich program here: sites.google.com/view/loresm…
And join us: aacl2020virtualconference.or…
@chstoneliu @surafelml @ffFlammie @alienelf
Surafel | ሱራፌል retweeted
Biased gender representations can be a matter of stylistic nuance. Can they be automatically rewritten to ensure more agency and power?
From my #EMNLP2020 highlights, "PowerTransformer: Unsupervised Controllable Revision for Biased Language Correction"
aclweb.org/anthology/2020.em…
Surafel | ሱራፌል retweeted
Masakhane is excited to present the keynote at WMT at #EMNLP2020 tomorrow! 🤩 We're on at 13h00 UCT. Join us for our talk, Q&A & a panel on low-resourcedness.
🌍✨Low-resourcedness Beyond Data✨🌍
workshop:
virtual.2020.emnlp.org/works…
paper:
aclweb.org/anthology/2020.fi…
Surafel | ሱራፌል retweeted
Working on speech translation?
Tired of the same old target languages? 🥱
New version of MuST-C now available!
📌 Extended with 6 new directions:
English➡️Arabic/Chinese/Czech/Persian/Turkish/Vietnamese
ict.fbk.eu/must-c/
#nlproc #machinetranslation #speechtranslation
Surafel | ሱራፌል retweeted
With overwhelming excitement that we can finally share with you "Participatory Research for Low-resourced Machine Translation: A Case Study in African Languages" to be published at ✨Findings of @emnlp2020 ✨💕🌍💪🏾
Preprint: arxiv.org/abs/2010.02353
/1
👇
REMINDER: October 11th is the deadline for the 2020 Amazon Research Awards. Call for proposal areas include:
1. AI for Information Security
2. Alexa Fairness in AI
3. AWS AI
4. AWS Automated Reasoning
5. Robotics
Apply here: amazon.science/research-awar…
#AmazonResearchAwards
Working on low-resource languages? don't miss this opportunity, highly recommend!
Get your proposals ready! ✍️
We’re excited to announce Lacuna Fund’s second RFP, focusing on datasets for language in sub-Saharan Africa. If you know somebody who’s #ClosingDataGaps to build better #NLProc #AI tools in African languages, please share! Learn more at lacunafund.org/apply
Surafel | ሱራፌል retweeted
We're overwhelmingly excited to announce our work✨ Participatory Research for Low-resourced Machine Translation: A Case Study in African Languages✨ is to be published at "Findings of @emnlp2020 " 🌍
Our paper boasts 50 authors from across African soil and beyond 🤩🌻💪🏿 /1
Surafel | ሱራፌል retweeted
LoResMT deadline extended to Sept. 18th. We are looking forward to your submissions! #MT #LoRes
sites.google.com/view/loresm…
A tip (or not) -- on reviewing am witnessing these days:
> reading a paper & making an immediate decision may not be the best of my decisions. As in most important decisions, am keeping the judgment for next morning :)
Surafel | ሱራፌል retweeted
Excited to share Lesan!
Check it out
lesan.ai/translate
መልካም አዲስ አመት ፳፻፲፫።
Happy Ethiopian New Year 2013.
#Lesan #Ethiopic #ልሳን
Apart from the impressive progress, key takeaways:
"
- poor performance on particular genres/domains
- conflating different dialects of a language
- producing overly literal translations
- poor performance on informal and spoken language
"
ai.googleblog.com/2020/06/re…
