Joe Melling
@joemelling.bsky.social
220 followers 420 following 15 posts
PhD candidate, studying cognitive science of (mis)perception/ philosophy of Active Inference @monash-m3cs.bsky.social (M3CS).
Posts Media Videos Starter Packs
Pinned
joemelling.bsky.social
Just realised Bluesky supports videos so here is our new visual illusion!

Fix your gaze on the grey dot. How many red dots do you see within the green frames? Many people consistently see two, despite only ever being shown one!

jov.arvojournals.org/article.aspx... w/ @renrutmailliw.bsky.social
Reposted by Joe Melling
plosbiology.org
It takes time for the #brain to process information, so how can we catch a flying ball? @renrutmailliw.bsky.social &co reveal a multi-stage #motion #extrapolation occurring in the #HumanBrain, shifting the represented position of moving objects closer to real time @plosbiology.org 🧪 plos.io/3Fm83Fc
Mapping the position of moving stimuli. The top three panels show the three events of interest: stimulus onset, stimulus offset, and stimulus reversal (left to right). The bottom three panels show group-level probabilistic spatio-temporal maps centered around these three events. Diagonal black lines mark the true position of the stimulus. Horizontal dashed lines mark the time of the event of interest (stimulus onset, offset, or reversal). Red indicates high probability regions and blue indicates low probability regions (‘position evidence’ gives the difference between the posterior probability and chance). Note: these maps were generated from recordings at posterior/occipital sites.
Reposted by Joe Melling
matthewod.bsky.social
Takeaways: musicians better in SJ tasks (but may be response bias) yet show > multisensory integration in a more objective RT task (but may be just better sustained attention, motivation, etc.). SJs show rapid recalibration but RTs don't, so RR probs not due to early sensory latency changes 9/10
Reposted by Joe Melling
matthewod.bsky.social
⚠️New paper! If you like multisensory temporal perception, this one has HEAPS for you. We looked at simultaneity perception, simple RTs, race model inequality, serial dependence (recalibration), modality shift costs, musicians, and more! Findings summarised below 1/10

psycnet.apa.org/fulltext/202...
APA PsycNet
psycnet.apa.org
joemelling.bsky.social
Ah, yes, that makes sense. An apt description, then. A lot of people also see the two dots as appearing semi-transparent (even in the conditions where they fully opaque, I mean), which might contribute to the effect you're describing.
joemelling.bsky.social
Hmm... I'm not sure what you mean exactly. Are you saying the frames double too? The common reported experience is that only the dots "split", there is no illusory effects w/r/t the frames.
joemelling.bsky.social
About 10-15% of people never saw the illusion in our experiment's viewing conditions so there may just be some intractable individual differences. I have always been very susceptible to this illusion so it's been hard to relate 😅
joemelling.bsky.social
Most people report seeing two distinct dots either side-by-side or with some diagonality - typically with the dot on the left being lower and the right being higher. People also sometimes reported seeing "oblongs" or "rectangles" which we assumed to mean there was partial separation.
joemelling.bsky.social
Just realised Bluesky supports videos so here is our new visual illusion!

Fix your gaze on the grey dot. How many red dots do you see within the green frames? Many people consistently see two, despite only ever being shown one!

jov.arvojournals.org/article.aspx... w/ @renrutmailliw.bsky.social
joemelling.bsky.social
Thanks for saying! We played around with initial demo *a lot* to try and maximise the effect. We found lots of individual difference, but most see it inconsistently. It's an open question as to what produces the inconsistency. We posit something attentional although there's prob. multiple factors.
joemelling.bsky.social
Hello Prof Miller. Bluesky didn't support uploading mp4 files when I originally posted, but it does now:
joemelling.bsky.social
Thanks to Will and Hinze for being great supervisors, and to the rest of the Timing Lab for their support, particularly @tvcottier.bsky.social for helping set up the testing room itself and Ben Lowe for code troubleshooting #BenWasHere [5/5]
joemelling.bsky.social
Thanks to Will and Hinze for being great supervisors, and to the rest of the Timing Lab for their support, particularly @tvcottier.bsky.social for helping set up the testing room itself and Ben Lowe for code troubleshooting #BenWasHere [5/5]
joemelling.bsky.social
This illusion has a number of unique properties when compared to other motion illusions, but we also argue, using a predictive processing framework, that it has important implications for how the brain constructs visual experience in space and time more generally. [4/5]
joemelling.bsky.social
In our illusion, we used two overlapping frames moving symmetrically and transparently. When a stimulus was flashed within the two frames as they both reversed direction, two concurrent mislocalisations can be perceived. [3/5]
joemelling.bsky.social
The illusion was inspired by Özkan et al.'s (2021) paradoxical stabilisation of moving frames, where the position of a flashed stimulus is misperceived due to the motion signals from a surrounding frame as it reverses direction. [2/5] www.pnas.org/doi/10.1073/...
Paradoxical stabilization of relative position in moving frames | PNAS
To capture where things are and what they are doing, the visual system may extract the position and motion of each object relative to its surroundi...
www.pnas.org
joemelling.bsky.social
Excited to share my first paper: a novel visual illusion discovered by my co-authors Will Turner and Hinze Hogendoorn which we call the "Split-Stimulus Effect", in which a single flashed stimulus is perceived to be in two different locations simultaneously jov.arvojournals.org/article.aspx... [1/5]
Concurrent perception of competing predictions: A “split-stimulus effect” | JOV | ARVO Journals
jov.arvojournals.org
joemelling.bsky.social
In our illusion, we used two overlapping frames moving symmetrically and transparently. When a stimulus was flashed within the two frames as they both reversed direction, two concurrent mislocalisations can be perceived. [3/5]
joemelling.bsky.social
The illusion was inspired by Özkan et al.'s (2021) paradoxical stabilisation of moving frames, where the position of a flashed stimulus is misperceived due to the motion signals from a surrounding frame as it reverses direction. www.pnas.org/doi/10.1073/... [2/5]
Paradoxical stabilization of relative position in moving frames | PNAS
To capture where things are and what they are doing, the visual system may extract the position and motion of each object relative to its surroundi...
www.pnas.org