Tom Wallis
@tsawallis.bsky.social
Vision scientist. Professor at the Centre for Cognitive Science, Technical University of Darmstadt, Germany. 🇦🇺🇩🇪. pronoun.is/he
https://www.psychologie.tu-darmstadt.de/perception
https://www.psychologie.tu-darmstadt.de/perception
I won't get to join you all at #ecvp2025 @ecvp.bsky.social this year because I'm on holiday in Australia... but my lab is there in force! Please do go say hi, and check out my students' and postdocs' amazing work!
@ben.graphics @lschmittwilken.bsky.social @lreining.bsky.social
@ben.graphics @lschmittwilken.bsky.social @lreining.bsky.social
August 25, 2025 at 10:56 AM
I won't get to join you all at #ecvp2025 @ecvp.bsky.social this year because I'm on holiday in Australia... but my lab is there in force! Please do go say hi, and check out my students' and postdocs' amazing work!
@ben.graphics @lschmittwilken.bsky.social @lreining.bsky.social
@ben.graphics @lschmittwilken.bsky.social @lreining.bsky.social
If I had to pick just two of these that you could implement now, it would be [1] use a consistent and clear directory structure for all projects in the group, and [2] did someone review your code? (and you walked them through it / answered questions they had).
April 16, 2025 at 3:08 PM
If I had to pick just two of these that you could implement now, it would be [1] use a consistent and clear directory structure for all projects in the group, and [2] did someone review your code? (and you walked them through it / answered questions they had).
I’m in beautiful (sunny!?) Scotland for the Scottish Vision Group meeting. Was nice to present some new work from the lab yesterday; looking forward to hearing about everyone’s work today. Thanks to @scienceanna.bsky.social and Alasdair Clarke for inviting me!
April 5, 2025 at 6:48 AM
I’m in beautiful (sunny!?) Scotland for the Scottish Vision Group meeting. Was nice to present some new work from the lab yesterday; looking forward to hearing about everyone’s work today. Thanks to @scienceanna.bsky.social and Alasdair Clarke for inviting me!
March 28, 2025 at 1:48 PM
I’ve not seen that exact case before, but David Regan’s book “Human perception of objects” has a bookmark that lets you do this:
November 29, 2024 at 4:33 PM
I’ve not seen that exact case before, but David Regan’s book “Human perception of objects” has a bookmark that lets you do this:
Verified just now in Python:
November 26, 2024 at 10:50 AM
Verified just now in Python:
First attempt at Pad Thai. Needed more sauce and less gluggy noodles.
November 25, 2024 at 6:53 AM
First attempt at Pad Thai. Needed more sauce and less gluggy noodles.
I think I make pretty good gluten-free pizzas.
November 16, 2024 at 10:47 PM
I think I make pretty good gluten-free pizzas.
This result is confirmed by a second experiment in which we also varied the level of smoothing. The level of smoothing has a higher influence on interpretability than the thresholding technique. However, what seems to make the most difference is the choice of template images.
September 30, 2024 at 10:30 AM
This result is confirmed by a second experiment in which we also varied the level of smoothing. The level of smoothing has a higher influence on interpretability than the thresholding technique. However, what seems to make the most difference is the choice of template images.
Although we found that the techniques generate physically-different Mooney images, identification performance and subjective ratings of identifiability were similar across the different techniques.
September 30, 2024 at 10:28 AM
Although we found that the techniques generate physically-different Mooney images, identification performance and subjective ratings of identifiability were similar across the different techniques.
The Mooney image generation process can be separated into image smoothing and image thresholding. We generated Mooney images using four different thresholding techniques, selected to represent various global thresholding methods, and parametrically varied the level of smoothing.
September 30, 2024 at 10:25 AM
The Mooney image generation process can be separated into image smoothing and image thresholding. We generated Mooney images using four different thresholding techniques, selected to represent various global thresholding methods, and parametrically varied the level of smoothing.
It seems that Bluesky doesn't handle transparency in a png image consistently. Here's the image again as a jpg in case you can't read the axes in the above plot.
September 6, 2024 at 5:55 AM
It seems that Bluesky doesn't handle transparency in a png image consistently. Here's the image again as a jpg in case you can't read the axes in the above plot.
5/ First, applying defensible transformations (rank order, log odds) to the proportion data shows no evidence that the detrimental effect of color removal is specific to Prakash patients. (Red is color, blue is greyscale images).
September 6, 2024 at 5:51 AM
5/ First, applying defensible transformations (rank order, log odds) to the proportion data shows no evidence that the detrimental effect of color removal is specific to Prakash patients. (Red is color, blue is greyscale images).
This is fun! Here's mine:
November 17, 2023 at 8:47 AM
This is fun! Here's mine:
