YouTube2Text

Ruihan Gao - 2026 Three Minute Thesis (3MT) Championship Presentation at CMU — Transcript

by Carnegie Mellon University Libraries · 437 words · 77 segments · language en · Watch on YouTube

Full transcript

  1. 0:01Our sixth presenter is Ruihan Gao from
  2. 0:03the School of Computer Science. Ruihan's
  3. 0:06title is Teaching AI the Sense of Touch.
  4. 0:11Raise your hand if you ever bought
  5. 0:12something online that looked great on
  6. 0:14the screen, but feels still totally
  7. 0:16different in your hands.
  8. 0:19Now, keep your hands up if that was the
  9. 0:21reason you sent it back.
  10. 0:23Exactly. Thank you. That happens because
  11. 0:26screens only show us how things look,
  12. 0:28but not how they feel like.
  13. 0:31Today, generative AI can create images,
  14. 0:34videos, 3D worlds in games, but they are
  15. 0:37missing something fundamental. Touch.
  16. 0:40As humans, we not only see the world, we
  17. 0:44feel the softness of fabric, the
  18. 0:46friction, the roughness of the wood, and
  19. 0:48also the coolness of glass.
  20. 0:51Touch can tell us something vision
  21. 0:53cannot.
  22. 0:55But, modern AI normally ignores it for
  23. 0:58two hard reasons.
  24. 1:00First, the scale gap. The camera sees
  25. 1:03the whole sweater, but the fingertip
  26. 1:05feels it thread by thread, almost 100
  27. 1:08times smaller.
  28. 1:10Second, unlike language, touch doesn't
  29. 1:13have a vocabulary. How do you quantify
  30. 1:15how a velvet feels like?
  31. 1:18My research solves both.
  32. 1:20We use a digital fingertip, a
  33. 1:22high-resolution sensor, and press it
  34. 1:24against object surface, and record the
  35. 1:27invisible heels and valleys into a
  36. 1:30texture map, and train the AI model to
  37. 1:33binds the big picture we see and the
  38. 1:35macroscopic details we feel
  39. 1:38into an integrated shared
  40. 1:40representation.
  41. 1:42The results, when you type smooth marble
  42. 1:45or rough linen, with the AI model can
  43. 1:48generate both it feels like and how it
  44. 1:51looks like.
  45. 1:53For you, that means an immersive life.
  46. 1:56Imagine feeling the weave of a sweater
  47. 1:58on your phone when you buy it.
  48. 2:00Or run your fingers across the stone
  49. 2:03wall in a video game.
  50. 2:05Beyond convenience, that's about
  51. 2:07accessibility.
  52. 2:09For blind and low vision individuals,
  53. 2:11touch isn't just a feature.
  54. 2:14It's their primary way of understanding
  55. 2:17the world.
  56. 2:18My work allows AI to generate tactile
  57. 2:21graphics, the 3D printed surfaces that
  58. 2:24translate images into the objects people
  59. 2:27can explore with their hands.
  60. 2:30Blind students, for the very first time,
  61. 2:32can see how a biological cell structured
  62. 2:35or touch a painting in the museum.
  63. 2:39We have demonstrated for rigid surfaces
  64. 2:41and fabric, and next we're modeling
  65. 2:43softness, so that medical medical
  66. 2:45students can play with the simulators of
  67. 2:48the soft skin without a
  68. 2:51you know, a simulator that finally feels
  69. 2:53real.
  70. 2:55The future of AI should not be limited
  71. 2:58to a flat glass screen. With us teaching
  72. 3:02the AI the sense of touch, we want the
  73. 3:04technology to be more sensory, more
  74. 3:07inclusive, and ultimately more human.
  75. 3:10Thank you.
  76. 3:11>> [applause]
  77. 3:13>> Thank you.

About this transcript

This page contains the full transcript of Ruihan Gao - 2026 Three Minute Thesis (3MT) Championship Presentation at CMU by Carnegie Mellon University Libraries, generated from the public captions YouTube serves with the video. The transcript has 437 words across 77 segments, with the original timestamps preserved so you can click any line to jump to that moment in the embedded player.

What you can do with it

Use the transcript to take notes, quote the speaker, build a study guide, generate a summary with ChatGPT or Claude via the YouTube Summary tool, or export it as a timed subtitle file with YouTube to SRT. You can also re-open it in the transcriber to translate the transcript into 100+ languages.

Free YouTube transcript tool

YouTube2Text is a free YouTube transcript generator — no signup, no daily limit. Paste any YouTube link and get the full transcript instantly, with timestamps, click-to-jump, translation to 100+ languages, AI prompts for ChatGPT, Claude, and Gemini, and exports to TXT, SRT, VTT, or Markdown.