Zubair Irshad
@mzubairirshad
Followers
1K
Following
1K
Media
73
Statuses
320
Research Scientist @ToyotaResearch | PhD in AI and DL @GeorgiaTech | Researching Large Behavioral Models | 3D Vision | Robotics
Silicon Valley, CA, USA
Joined June 2014
Happy to announce I successfully defended my PhD dissertation yesterday! š. Really grateful for all who helped along in this journey, with special thanks to my advisor @zsoltkira for teaching me how to do great research and guidance at every step along the way!
4
6
108
RT @kushalk_: Teleoperation is slow, expensive, and difficult to scale. So how can we train our robots instead?. Introducing X-Sim: a real-ā¦.
0
38
0
RT @KarlPertsch: Weāre releasing the RoboArena today!š¤š¦¾. Fair & scalable evaluation is a major bottleneck for research on generalist policiā¦.
0
80
0
RT @robo_kat: How can we achieve both common sense understanding that can deal with varying levels of ambiguity in language and dextrous maā¦.
0
5
0
RT @zsoltkira: @CVPR next week will be an exciting one! Check out our work below on VLMs, VLAs, and 3D for robotics (including the first 3Dā¦.
0
3
0
RT @jiading_fang: Ever want to reconstruct and animate everyday articulated objects with no 3D scans or category priors?. šIntroducing SplAā¦.
0
1
0
Interested in collecting robot training data without robots in the loop? š¦¾ Check out this cool new approach that uses a single mobile device scan and a human demo video to generate diverse data for training diffusion and VLA manipulation policies. š. Great work by @letian_fu.
Tired of teleoperating your robots?.We built a way to scale robot datasets without teleop, dynamic simulation, or even robot hardware. Just one smartphone scan + one human hand demo video ā thousands of diverse robot trajectories. Trainable by diffusion policy and VLA models
0
2
13
RT @letian_fu: Tired of teleoperating your robots?.We built a way to scale robot datasets without teleop, dynamic simulation, or even robotā¦.
0
75
0
RT @Almorgand: FastMap: Revisiting Dense and Scalable Structure from Motion. TL;DR: 2 orders of magnitude faster than GLOMAP; many GPU implā¦.
0
19
0
RT @janusch_patas: FastMap: Revisiting Dense and Scalable Structure from Motion. "FASTMAP, a redesigned SfM framework, achieves fast, high-ā¦.
0
46
0
RT @zhenjun_zhao: FastMap: Revisiting Dense and Scalable Structure from Motion. Jiahao Li, @__whc__, @mzubairirshad, @vslevic, Matthew R. Wā¦.
0
21
0
@vitorguizilini @SashaKhazatsky @KarlPertsch Shoutout to the authors of the wonderful papers i.e. CtRNet-X, DUSt3R, Segment Anything, CLIP and Pytorch3D and for open-sourcing their codebase to advance science and make this effort happen!. Please check these works out if you havenāt already!.
0
0
2
We have released our improved extrinsics. Try it out now at and read more details about it in the updated DROID paper at This was a fun collaboration with @vitorguizilini @SashaKhazatsky and @KarlPertsch!.
1
0
0
Automatic calibration robotics at large-scale is challenging and while successful, our pipeline has some limitations:. ā¢ CtRNet-X was trained on the Franka Panda robot. Its performance i.e. zero-shot generalizability on other robot types remains to be seen. ā¢ DUSt3R, while.
1
0
1
Similarly, we plot the distribution of number of matched points and cumulative curve after 3ļøā£, helping to identify the top quantile of well-calibrated camera pairs within each lab. 4/n
1
0
0
Automatically calibrating a large-scale dataset is challenging. We provide quality assessment metrics with all three stages with the flexibility to narrow out bounds for downstream tasks as needed. 1ļøā£ and 2ļøā£ quality assessment metrics shows distribution of IOUs and
1
0
0
Below we show the Camera-to-Camera transformations, post-calibration improves the alignment of obtained pointclouds!. 2/n
1
1
4
Introducing āØPosed DROIDāØ, results of our efforts at automatic post-hoc calibration of a large-scale robotics manipulation dataset. We provide:.š¤ ~36k calibrated episodes with good quality extrinsic calibration.š¦¾ ~24k calibrated multi-view episodes with good-quality
3
27
188