Weihan Luo
@luo_weihan
Followers
57
Following
738
Media
6
Statuses
37
RT @sherwinbahmani: π’ Excited to share our new work: AC3D: Analyzing and Improving 3D Camera Control in Video Diffusion Transformers. httpsβ¦.
0
24
0
Check out some early work that inspired us:.-NLOS-NeuS by Fujimura et al. -Neuralangelo by @mli0603 et al. -RegNeRF by @Mi_Niemeyer et al. -RawNeRF @BenMildenhall et al.
0
0
1
Overall, we achieve SOTA results on both the simulated and captured datasets across various photon levels, with as few as an average of 10 photons per pixel. 3/.(more results in the paper)
1
0
2
To address these issues, we model Transientangelo as a surface-based representation with the Eikonal constraint. We additionally supervise the network from unseen viewpoints, by constraining the weight variance and adding a supervision on the integrated transients. 2/
1
0
1
Prior work TransientNeRF, which optimizes a lidar-based volumetric representation, recovers sub-par meshes that are often accompanied by floating artifacts. Furthermore, it struggles in low-photon scenarios. 1/.Left: TransientNeRF, right: our method.(more videos on the website)
1
0
1
π£ Excited to share a project Iβve been working on:.βTransientangelo: Few-Viewpoint Surface Reconstruction Using Single-Photon Lidarβ. We render more accurate surfaces compared to prior baselines when trained on (2-5) input single-photon lidar views. web:Β
2
14
49
RT @zhenjun_zhao: Transientangelo: Few-Viewpoint Surface Reconstruction Using Single-Photon Lidar. Weihan Luo, @anagh_malik, @DaveLindellβ¦.
0
4
0
RT @_sam_sinha_: I am SO excited to share what me and the amazing @LumaLabsAI team has been working on for the past many months!! π₯°. A statβ¦.
0
11
0
RT @Guang_LI0: ECCV 2024 has officially accepted our proposal to host the 1st Dataset Distillation Challenge. This event will be co-hostedβ¦.
0
12
0
RT @anagh_malik: π’π’π’ A pulse of light takes ~3ns to pass through a Coke bottleβ100 million times less than it takes you to blink. Our workβ¦.
0
40
0
RT @sherwinbahmani: Happy to share our new work π₯³.TC4D: Trajectory-Conditioned Text-to-4D Generation. Project page: .
0
17
0
RT @skirano: I used Claude 3 to unredact this part from the OpenAI emails. What's wild is that they used a "per word" redaction, meaning eβ¦.
0
210
0
RT @anagh_malik: π’π’π’ We will be presenting Transient NeRF at #NeurIPS2023 next week!. Come by our poster session on Thursday at 10.45AM CSTβ¦.
0
4
0
RT @taiyasaki: π’π’π’ thrilled to announce "ππ-ππ²: πππ±π-ππ¨-ππ πππ§ππ«πππ’π¨π§ ππ¬π’π§π ππ²ππ«π’π πππ¨π«π ππ’π¬ππ’π₯π₯πππ’π¨π§ πππ¦π©π₯π’π§π ". Wβ¦.
0
76
0
