Challenges
Abstract
We present the HANDS23 challenge for hand pose estimation based on AssemblyHands and ARCTIC. In line with this year’s theme, the challenge consists of two tasks and focuses on hand pose estimation on occlusion and interaction scenarios.
Our first task focuses on egocentric 3D hand pose estimation from a single-view image, which has been less explored due to the limited benchmarks for this task but has great potential and attention for enabling AR/VR applications. We use the recently introduced AssemblyHands. It provides mutli-view captured videos of hand-object interaction during assembling and disassembling of toy-vehicles. In particular, it provides both static and egocentric recordings, and auxiliary cues like action, object, or context information for hand pose estimation.
Our second task focuses on consistent motion reconstruction based on ARCTIC. It is a dataset of hands dexterously manipulating articulated objects and contains videos from 8x 3rd-person views and 1x egocentric view. Besides accurate ground-truth 3D hand and object meshes, it also provides detailed contact information between the hands and objects during manipulation.
Winners and prizes will be announced and awarded during the workshop.
Please visit the challenge pages for more details.
Challenges Page
Participation
We follow the rules of previous challenges and more details can be found in the challenge page.
To participate the challenge, please fill the form and accept the terms and conditions.
The evaluation server has opened since August 27 for AssemblyHands and September 9 for ARCTIC (Updated September 9).
The challenge start on June 15 2023 and end on September 15 2023 for AssemblyHands and September 23 2023 for ARCTIC (Updated September 9).
We will contact participants via emails once we have an important update.
Submissions details are provided in the challenge page.
Each team must register under only one user id/email id. Teams found to be registered under multiple IDs will be disqualified.
In order for participants to be eligible for competition prizes and be included in the official rankings (to be presented during the workshop and subsequent publications), information about their submission must be provided to organizers. Information may include, but not limited to, details on their method, synthetic and real data use, architecture and training details. Check previous challenge publication to have an idea of the information needed.
Winning methods may be asked to provide their source code to reproduce their results under strict confidentiality rules if requested by organizers/other participants.
For each submission, participants must keep the parameters of their method constant across all testing data for a given task.
We invite all winners to advertise their methods via any combination of the following: 2-3 page technical report, winner talk, and poster presentation.