[Adapt] [Seminar] Generative Imagination Elevates Machine Translation

Kenny Zhu kzhu at cs.sjtu.edu.cn
Wed Nov 23 10:14:56 CST 2022


Because some adapters are locked in their dorm rooms today. We will have a
mixed mode. Leo, can you send us the tencent meet id as well so they can
join?

Kenny

-----Original Message-----
From: Adapt <adapt-bounces at cs.sjtu.edu.cn> On Behalf Of 罗嘉鸣
Sent: 2022年11月23日 10:09
To: adapt <adapt at cs.sjtu.edu.cn>
Subject: [Adapt] [Seminar] Generative Imagination Elevates Machine
Translation

Hi Adapters,

Multimodal machine translation (MMT) has been a hot topic in the past years.
However, due to the difficulty of collecting sufficient image-text pairs,
there are still some limitations of MMT in application. Today I will
introduce the paper: Generative Imagination Elevates Machine Translation,
which was accepted by NAACL 2021. This paper proposed ImagiT, which
generates images from raw texts and used MMT model to solve the problem.

Wish you would like it!

Time: Wed 4:00pm
Venue: SEIEE 3-414

Best regards,
Leo
_______________________________________________
Adapt mailing list
Adapt at cs.sjtu.edu.cn
http://cs.sjtu.edu.cn/mailman/listinfo/adapt



More information about the Adapt mailing list