Shubham Tulsiani's banner
Shubham Tulsiani's profile picture

Shubham Tulsiani

@shubhtuls • 3,429 subscribers

Assistant Professor in the Robotics Institute, Carnegie Mellon University I want to build perception systems that can understand the physical world

Shorts

[1/6] What representation comes to mind when you think of a ‘camera’? Perhaps an extrinsic + intrinsic matrix? In our ICLR (oral) paper, we instead infer a distributed representation where each pixel is associated with a ray, and show SoTA results for few-view pose estimation.

[1/6] What representation comes to mind when you think of a ‘camera’? Perhaps an extrinsic + intrinsic matrix? In our ICLR (oral) paper, we instead infer a distributed representation where each pixel is associated with a ray, and show SoTA results for few-view pose estimation.

141,781 次观看

[1/6] Our #CVPR2025 paper “DiffusionSfM” extends our RayDiffusion framework — inferring both geometry and cameras via diffusing pixelwise ray origins and endpoints.

[1/6] Our #CVPR2025 paper “DiffusionSfM” extends our RayDiffusion framework — inferring both geometry and cameras via diffusing pixelwise ray origins and endpoints.

41,433 次观看

Videos

LIVE
1.2k

Anya Rossi

sweetdream.ai

SweetDream.aiSponsored•Livecam
Streaming Now

Watch Anya Live

Anya is streaming live right now! Join her private show and enjoy exclusive content.

HD live stream
Exclusive private shows
1.2k viewers online

Free preview available • Premium content

[1/N] Rotary Position Embeddings (RoPE) are ubiquitous across transformers that process tokens from 1D, 2D, or 3D grids e.g. language, images, or videos. Our RayRoPE formulation extends these to multi-view transformers. Paper and code:
0:41
shubhtuls's profile picture

[1/N] Rotary Position Embeddings (RoPE) are ubiquitous across transformers that process tokens from 1D, 2D, or 3D grids e.g. language, images, or videos. Our RayRoPE formulation extends these to multi-view transformers. Paper and code:

Shubham Tulsiani

55,859 次观看 • 4 个月前

[1/N] We present a plug-and-play mechanism to controllably steer inference of any diffusion/flow model towards a sharper or flatter sampling distribution, resulting in improvements across domains e.g. text-to-image (10% FID reduction), protein generation (improved designability).
0:51
shubhtuls's profile picture

[1/N] We present a plug-and-play mechanism to controllably steer inference of any diffusion/flow model towards a sharper or flatter sampling distribution, resulting in improvements across domains e.g. text-to-image (10% FID reduction), protein generation (improved designability).

Shubham Tulsiani

60,777 次观看 • 8 个月前

[1/N] Current visual geometry prediction models primarily rely on labeled 3D data. Our CVPR26 paper, Flow3r, allows additionally leveraging unlabeled videos (using flow supervision) for scalable visual geometry learning, enabling accurate multi-view 3D reconstruction in-the-wild.
0:31
shubhtuls's profile picture

[1/N] Current visual geometry prediction models primarily rely on labeled 3D data. Our CVPR26 paper, Flow3r, allows additionally leveraging unlabeled videos (using flow supervision) for scalable visual geometry learning, enabling accurate multi-view 3D reconstruction in-the-wild.

Shubham Tulsiani

15,974 次观看 • 3 个月前

没有更多内容可加载