wimpouw
banner
wimpouw.bsky.social
wimpouw
@wimpouw.bsky.social
Working on dynamic signs and signals in communication at Tilburg University at the Department of Computational Cognitive Science. Interested in language & movement science, complex systems & 4E, Open Science (envisionbox.org) & other things (wimpouw.com)
This shows that energy investment does not translate into communicative success, but rather the temporal dynamics of invested energy reflects the communicative work performed; a process that involves exploration-exploitation (we think).

Also with Tianyi Li, Marlou Rasenberg, Lotte Eijk, Ivan Toni
November 11, 2025 at 4:49 PM
Our paper @sarabogels.bsky.social covering our pre-registered multi-year research is now finally out in Cognition. We show that in conversations people reduce their multimodal signals non-linearly; the steeper this non-linear drop-off the more communicative success.

www.wimpouw.com/files/Bogels...
November 11, 2025 at 4:49 PM
At this workshop today about where hand gestures get their rhythm from - participants can join online.

sfb1252.uni-koeln.de/veranstaltun...
October 7, 2025 at 7:01 AM
So publishers, like Wiley, are now putting ads on their websites? Did I only notice now? What the hell
October 3, 2025 at 9:07 AM
mmmconflab.ewi.tudelft.nl

Check out this workshop coming up 17th of July in Rotterdam, still looking for participants!
July 14, 2025 at 4:19 PM
Our @dondersinst.bsky.social funded collaboration project EnvisionBoxBaby is in full swing. Leveraging object detection for eye tracking-object dynamics.

With @babajideowoyele.bsky.social @Davide Ahmar, @sabinehunnius.bsky.social @carorowland.bsky.social @yayun23.bsky.social
July 8, 2025 at 9:30 AM
We are looking for Dutch participants in Nijmegen who would like to do a study. We are quite desperate so thought to reach out here too. Perhaps for those Dutchies who are visiting @isgs2025.bsky.social
July 3, 2025 at 3:32 PM
In preparation of teaching and supervision coming years I have created a webpage that provides a bit of a landing page for students to see what me and my colleagues are up to, to help carve out projects: www.wimpouw.com/collaborate....
July 3, 2025 at 11:11 AM
Paper led by the amazing Rui Liu (www.ugent.be/pp/en/resear...), shows how temporal rather than spatial kinematic features are modulated in communicative vs. instrumental movements (but in opposite ways for signers versus speakers) - now accepted for publication in Open Mind.
June 24, 2025 at 4:34 PM
We are looking forward to our next EnvisionBox summerschool next week at the University of Amsterdam. We are also very pleased with the poster (with new logo for envisionbox)!
@jamestrujillo.bsky.social
@babajideowoyele.bsky.social @skadava.bsky.social
June 20, 2025 at 11:24 AM
Postprint: osf.io/987fm_v1 To appear in Proceedings of Cog Sci 2025

DIMS Dashboard for Exploring Dynamic Interactions and Multimodal Signals.

The interdisciplinary @graceqmiao.bsky.social in the lead here! Developing a dynamic dashboard for a quali-quanti social neuroscience research cycle!
May 12, 2025 at 1:00 PM
Applications open! forms.gle/LnNSEyivsADr...
April 24, 2025 at 4:35 PM
Its full of multimodal signal processing and analyses - continuous and peak-based. Methods & results (and prereg) were fully written in Rmarkdow.

This was a lot of work, and were happy it now has a peer-reviewed stamp. wimpouw.github.io/kineticsvoice/
March 25, 2025 at 8:13 AM
This was a key step to show that voice and upper limb movement couple through dynamics (not just kinematics), and underlines how open the vocal system is to interactions from the whole body (re Pouw & Fuchs, wimpouw.com/files/PouwFu...).
March 25, 2025 at 8:13 AM
Update: Our biomechanics EMG/Posture study on voice and upper limb kinetics has been accepted at Proceedings B. We show that moving the arms (while manipulating mass) recruits (postural) muscles that interact with the voice through respiratory interactions. Postprint: wimpouw.com/files/Pouwet...
March 25, 2025 at 8:13 AM
We further provides tests of performance and future outlook. Also see for open source code and training data the paper's links (osf.io/preprints/ps...). Much more to come!

For a tutorial of using this on your own videos see here: envisionbox.org/embedded_Usi...
February 27, 2025 at 8:43 AM
The package automatically detects gestures using a CNN classifier generating gesture segments and ELAN files, further performing gesture kinematic feature and DTW analysis, as to then visualizes everything interactively with a dashboard app (see paper for the server version).
osf.io/preprints/ps...
February 27, 2025 at 8:43 AM
Update re pypi package envisionHGdetector! Our preprint "EnvisionHGdetector: A Computational Framework for Co-Speech Gesture Detection, Kinematic Analysis, and Interactive Visualization" is now out! osf.io/preprints/ps...

@jamestrujillo.bsky.social @gdemelo.bsky.social @babajideowoyele.bsky.social
February 27, 2025 at 8:43 AM
The email signature banner from from WO in Actie for the strike the 13th of March
February 18, 2025 at 8:43 AM
We (with Annika Schiefner) put a new EnvisionBox module online for creating stimuli where one wants to dynamically block the signal potential of different facial gestures, mouthings, or articulatory information. Maybe helpful for some!

envisionbox.org/embedded_Sti...

github.com/WimPouw/Auto...
February 7, 2025 at 2:19 PM
23-27June: The EnvisionBox team @jamestrujillo.bsky.social @babajideowoyele.bsky.social is organizing a summer school on social signal processing again this year! In Amsterdam - hosted by @jamestrujillo.bsky.social

Well send out an official sign-up announcement via the envisionbox newsletter soon.
January 31, 2025 at 11:55 AM
working gif of example output
December 19, 2024 at 9:57 AM
for training your own model, here is a runthrough from beginning to end based on open source pipeline NoddingPigeon: envisionbox.org/embedded_Hum...
December 19, 2024 at 8:05 AM
Made my first PyPi package "EnvisionHGdetector" (version 0.0.3), a simple skeleton-based hand gesture recognizer using a convolutional neural network trained on three fully open datasets (SAGA, TEDM3D, Zhubo).

Demo and info: envisionbox.org/embedded_Usi...
Package:
pypi.org/project/envi...
December 19, 2024 at 7:57 AM
dont do it
December 4, 2024 at 8:16 AM