Amirhossein Tebbifakhr

[Amirhossein Tebbifakhr @AT_Amir]

So much gratitude for everyone who was part of this journey!🙏

Special thanks to my supervisors @Turchi_Marco & @negri_teo, and @fbk_mt colleagues.

#PhDone

[MT Group at FBK @fbk_mt]

👏👏👏 Congratulations, Doctor @at_amir!

Many thanks to his committee members Josef van Genabith, Marco Maggini and @MauroCettolo from proud supervisors @Turchi_Marco @negri_teo 😁

#nlproc #phdlife #phdtimes #PhDone

[MT Group at FBK @fbk_mt]

Our student @at_amir ready to discuss his PhD research on machine translation for machines. Good luck for your viva! #nlproc #machinetranslation #phdviva @Turchi_Marco @negri_teo

[MT Group at FBK @fbk_mt]

Un asterisco sul genere - Consulenza Linguistica - Accademia della Crusca https://t.co/u7W10BKBEp

[MT Group at FBK @fbk_mt]

Come and join our team! We offer a fully funded 3-year PhD position:

Flexibility and Robustness in Speech Translation
https://t.co/9FnHbfiETW

Full details for application: https://t.co/ZagIUM3mMQ

Deadline: April 15, 2021

@UniTrento @FBK_research #NLProc

[Alina Karakanta @karakanta]

Starting our #EAMT sponsored project:

Towards a methodology
for evaluating automatic subtitling

We make the first steps towards a methodology for evaluating multiple aspects of quality in automatically generated subtitles.

Stay tuned!🥳
@fbk_mt @EAMTee

[Amirhossein Tebbifakhr @AT_Amir]

What are the challenges NLP tools facing to leverage user logs? Here is a comprehensive overview of Reinforcement Learning challenges in offline setting:

https://t.co/dTgiO2ChMH

@fbk_mt

[Alina Karakanta @karakanta]

Presenting our #coling2020 paper https://t.co/zALfj9oE8v

The Two Shades of Dubbing
in Neural Machine Translation

at poster session16, Dec 10th @ 16:00 CET
@karakanta
, Supratik Bhattacharya, Shravan Nayak, @realTimoBaumann
, @negri_teo
, @Turchi_Marco

@fbk_mt

[Marco Gaido @mgaido91]

Our paper on "Breeding Gender-aware Direct Speech Translation Systems" will be presented on Thursday 10th at 15:30-16:00 CEST.
@BeatriceSavoldi @fbk_mt @negri_teo @luisabentivogli @Turchi_Marco

[BeatriceSavoldi @BeatriceSavoldi]

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"
https://t.co/qrw1BT72xN

[Amirhossein Tebbifakhr @AT_Amir]

@di_tri_nel Do you have any plan to make public the data?

[Amirhossein Tebbifakhr @AT_Amir]

Another application of Reinforcement Learning in Machine Translation: https://t.co/OVHm2jR2NX

Lexical cohesion & coherence in document-level translation can be optimized by RL.

by: @ijauregi, N. Esmaili, @haffari & M. Piccardi

@fbk_mt

[MT Group at FBK @fbk_mt]

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

https://t.co/8EhWUA3pI6

#nlproc #machinetranslation #speechtranslation

[Marco Gaido @mgaido91]

Our paper on "Contextualized Translation of Automatically Segmented Audio" has been accepted to #Interspeech2020. See you virtually there!
@mdigangiPA @fbk_mt

[Marco Gaido @mgaido91]

Phoneme-based semi-unsupervised reduction of speech data has been shown to be helpful for some speech tasks at #ICASSP2020. Will this become a standard approach for other tasks such as speech translation? https://t.co/OHacqbCIqt

@fbk_mt

[Surafel | ሱራፌል @surafelml]

In multilingual and zero-shot translation tasks @BZhangGo et al. work is very interesting, showing improvements for

- low-resource translation with massive model capacity

+ zero-shot translation with a language-aware model component and back-translation

[Amirhossein Tebbifakhr @AT_Amir]

Interesting approach if you are thinking to leverage BERT in Neural Machine Translation.

https://t.co/BsH0EENfwf

@fbk_mt

[Amirhossein Tebbifakhr @AT_Amir]

@TheDemoVersion آموزشش رو میذاری؟

[Alina Karakanta @karakanta]

More #NMT works relevant for #subtitling! Here a nice overview of methods for translating with length constraints https://t.co/v3Y5XyOdoH
@fbk_mt

[Amirhossein Tebbifakhr @AT_Amir]

@jasmine_ici @amiiigh

[Amirhossein Tebbifakhr @AT_Amir]

Our paper: "Automatic Translation for Multiple NLP tasks: a Multi-task Approach to Machine-oriented NMT Adaptation" has been accepted at #EAMT2020
@negri_teo @Turchi_Marco

@fbk_mt