Upgrade to Pro — share decks privately, control downloads, hide ads and more …

Words or Vision: Do Vision-Language Models Have...

Avatar for Daiki Shiono Daiki Shiono
August 24, 2025
0
11

Words or Vision: Do Vision-Language Models Have Blind Faith in Text? (CVPR2025)

第17回最先端NLP勉強会 (SNLP2025) 発表資料

Avatar for Daiki Shiono

Daiki Shiono

August 24, 2025
Tweet

More Decks by Daiki Shiono

See All by Daiki Shiono
Large Vision Language Model (LVLM) に関する最新知見まとめ (Part 1)
Avatar for Daiki Shiono onely7
27
7.5k

Featured

See All Featured
Refactoring Trust on Your Teams (GOTO; Chicago 2020)
Avatar for Rebecca Miller-Webster rmw
34
3.1k
ピンチをチャンスに：未来をつくるプロダクトロードマップ #pmconf2020
Avatar for Aki / @LoveIdahoBurger aki_iinuma
126
53k
I Don’t Have Time: Getting Over the Fear to Launch Your Podcast
Avatar for Joe Casabona jcasabona
33
2.4k
Typedesign – Prime Four
Avatar for Hannes Fritz hannesfritz
42
2.8k
CoffeeScript is Beautiful & I Never Want to Write Plain JavaScript Again
Avatar for Sam Stephenson sstephenson
161
15k
How GitHub (no longer) Works
Avatar for Zach Holman holman
315
140k
How STYLIGHT went responsive
Avatar for nonsquared nonsquared
100
5.7k
Git: the NoSQL Database
Avatar for Brandon Keepers bkeepers
PRO
431
65k
What's in a price? How to price your products and services
Avatar for Michael Herold michaelherold
246
12k
Exploring the Power of Turbo Streams & Action Cable | RailsConf2023
Avatar for Kevin Liebholz kevinliebholz
34
6k
Building an army of robots
Avatar for Kyle Neath kneath
306
46k
Mobile First: as difficult as doing things right
Avatar for Swwweet swwweet
223
9.9k

Transcript

  1. Ԙ໺େً %BJLJ4IJPOP % (SBEVBUF4DIPPMPG*OGPSNBUJPO4DJFODFT 5PIPLV6OJWFSTJUZ !4/-1 "JMJO%FOH 5SJ$BP ;IJSVJ $IFO

    #SZBO)PPJ $713 1PTUFS <IUUQTDWQSUIFDWGDPNWJSUVBMQPTUFS>