Some recent interesting work on hand tracking and pose estimation that I liked. Creating a thread 🧵
MEgATrack: Monochrome Egocentric Articulated Hand Tracking for Virtual Reality…
The Phong Surface: Efficient 3D Model Fitting using Lifted Optimization
GRAB: A Dataset of Whole-Body Human Grasping of Objects
ContactPose: A Dataset of Grasps with Object Contact and Hand Pose
Weakly-Supervised Mesh-Convolutional Hand Reconstruction in the Wild…

Project website:
FingerTrak Continuous 3D Hand Pose Tracking by Deep Learning Hand

HOnnotate: A method for 3D Annotation of Hand and Object Poses
HandVoxNet: Deep Voxel-Based Network for 3D Hand Shape and Pose Estimation from a Single Depth Map…
Weakly Supervised 3D Hand Pose Estimation via Biomechanical Constraints

• • •

Missing some Tweet in this thread? You can try to force a refresh

Keep Current with Ankur Handa

Ankur Handa Profile picture

Stay in touch and get notified when new unrolls are available from this author!

Read all threads

This Thread may be Removed Anytime!


Twitter may remove this content at anytime! Save it as PDF for later use!

Try unrolling a thread yourself!

how to unroll video
  1. Follow @ThreadReaderApp to mention us!

  2. From a Twitter thread mention us with a keyword "unroll"
@threadreaderapp unroll

Practice here first or read more on our help page!

More from @ankurhandos

30 Aug
Interesting SIGGRAPH courses that I liked

1. Physics-Based Differentiable Rendering - A Comprehensive Introduction…
Understanding AR inside and out…
Virtual Hands in VR: Motion Capture, Synthesis, and Perception…
Read 4 tweets
19 May
GrokNet, a universal computer vision system designed for shopping is trained on **7 datasets and 83 loss functions**.…
Also... "we adjust the batch sizes and loss weights, using more images per batch and higher loss weights for the challenging tasks. We also use weakly supervised learning to automatically generate additional training data, further improving accuracy."
progress in segmentation over the years.
Read 5 tweets
26 Mar
I have been playing with my 2D toy example of NeRF ( that implemented to understand the role of positional encoding (PE).


left is the dataset image, middle is results with PE and right is without it. It really helps.
This is the summary of my experiment
This figure in the paper really intrigued me so I set up a very toy 2D example to understand this. The difference in the results is significant.
Read 11 tweets

Did Thread Reader help you today?

Support us! We are indie developers!

This site is made by just two indie developers on a laptop doing marketing, support and development! Read more about the story.

Become a Premium Member ($3/month or $30/year) and get exclusive features!

Become Premium

Too expensive? Make a small donation by buying us coffee ($5) or help with server cost ($10)

Donate via Paypal Become our Patreon

Thank you for your support!

Follow Us on Twitter!