Identification of movie encoding neurons enables movie recognition AI DOI Creative Commons
Masaki Hiramoto, Hollis T. Cline

Proceedings of the National Academy of Sciences, Journal Year: 2024, Volume and Issue: 121(48)

Published: Nov. 19, 2024

Natural visual scenes are dominated by spatiotemporal image dynamics, but how the system integrates “movie” information over time is unclear. We characterized optic tectal neuronal receptive fields using sparse noise stimuli and reverse correlation analysis. Neurons recognized movies of ~200-600 ms durations with defined start stop stimuli. Movie from to responses were tuned sensory experience though a hierarchical algorithm. encoded families sequences following trigonometric functions. Spike sequence flow suggest that repetitive circuit motifs underlie movie detection. Principles frog topographic retinotectal plasticity cortical simple cells employed in machine learning networks for static recognition, suggesting discoveries principles encoding brain, such as duration encoded, may benefit recognition technology. built trained network mimicked neural encoders. The network, named MovieNet, outperformed current classifying natural scenes, while reducing data size steps complete classification task. This study reveals brain demonstrates brain-based processing enable efficient learning.

Language: Английский

Optimization Extraction Method and Physicochemical Properties of <i>Citrus Sinensis L</i>. Peel Essential Oil for Enhancing Food Sustainability DOI

Wen Xia Ling Felicia,

Kobun Rovina,

Nur Aqilah Md Nasir

et al.

Published: Jan. 1, 2024

Download This Paper Open PDF in Browser Add to My Library Share: Permalink Using these links will ensure access this page indefinitely Copy URL DOI

Language: Английский

Citations

0
String-pulling by the common marmoset DOI Creative Commons

Mathilde Bertrand,

Michael Karkuszewski,

Rhonda Kersten

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2024, Volume and Issue: unknown

Published: Oct. 26, 2024

Abstract Coordinated hand movements used to grasp and manipulate objects are crucial for many daily activities, such as tying shoelaces or opening jars. Recently, the string-pulling task – which involves cyclically reaching, grasping pulling a string has been study coordinated in rodents humans. Here we characterize how adult common marmosets perform describe changes performance across lifespan. Marmosets (n=15, 7 female) performed food reward using an instrumented apparatus attached their home-cage. Movement kinematics were acquired markerless video tracking assessed individual bimanual coordination standard metrics. oriented gaze towards above hands readily regardless of sex age. The required very little training animals routinely engaged multiple trials per session despite not being under water control. All showed consistent speed similar Adult exhibited clear effect, performing straighter faster with right showing idiosyncratic preference according traditional retrieval assay. Hand effects also evident younger but seemed attenuated older animals. In terms coordination, all demonstrated alternating movement pattern vertical positions. Two two patterns even after substantial experience. general, higher variability than did adults.

Language: Английский

Citations

0
Late gene therapy limits the restoration of retinal function in a mouse model of retinitis pigmentosa DOI Creative Commons
Miranda L. Scalabrino, Mishek Thapa, Tian Wang

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: April 8, 2023

Retinitis pigmentosa is an inherited photoreceptor degeneration that begins with rod loss followed by cone and eventual blindness. Gene therapies are being developed, but it unknown how retinal function depends on the time of intervention. To uncover this dependence, we utilized a mouse model retinitis capable artificial genetic rescue. This enables benchmark best-case gene therapy removing variables complicate ability to answer vital question. Complete rescue was performed at 25%, 50%, 70% (early, mid late, respectively). Early treatment restored near wild-type levels, specifically sensitivity signal fidelity ganglion cells (RGCs), 'output' neurons retina. However, some anatomical defects persisted. Late retinas exhibited continued, albeit slowed, among RGCs, as well persistent gliosis. We conclude replacement delivered after 50% unlikely restore visual normal. critical information for administering vision.

Language: Английский

Citations

1
Establishing an AI-based evaluation system that quantifies social/pathophysiological behaviors of common marmosets DOI Creative Commons
Takaaki Kaneko, Jumpei Matsumoto,

Wanyi Lu

et al.

bioRxiv (Cold Spring Harbor Laboratory), Journal Year: 2023, Volume and Issue: unknown

Published: Oct. 17, 2023

Abstract Nonhuman primates (NHPs) are indispensable animal models by virtue of the continuity behavioral repertoires across primates, including humans. However, assessment at laboratory level has so far been limited. By applying multiple deep neural networks trained with large-scale datasets, we established an evaluation system that could reconstruct and estimate three-dimensional (3D) poses common marmosets, a small NHP is suitable for analyzing complex natural behaviors in setups. We further developed downstream analytic methodologies to quantify variety parameters beyond simple motion kinematics, such as social interactions internal state behind actions, obtained solely from 3D pose data. Moreover, fully unsupervised approach enabled us detect progressively-appearing symptomatic over year Parkinson’s disease model. The high-throughput versatile nature our pipeline will open new avenue neuroscience research dealing big-data analyses social/pathophysiological NHPs.

Language: Английский

Citations

1
AI-Assisted 3D Analysis of Grasping and Reaching Behavior of Squirrel Monkeys During Recovery from Cervical Spinal Cord Injury DOI Creative Commons

Daniela Hernandez Duque,

Pai‐Feng Yang, John C. Gore

et al.

Behavioural Brain Research, Journal Year: 2024, Volume and Issue: unknown, P. 115265 - 115265

Published: Sept. 1, 2024

Language: Английский

Citations

0
Identification of movie encoding neurons enables movie recognition AI DOI Creative Commons
Masaki Hiramoto, Hollis T. Cline

Proceedings of the National Academy of Sciences, Journal Year: 2024, Volume and Issue: 121(48)

Published: Nov. 19, 2024

Natural visual scenes are dominated by spatiotemporal image dynamics, but how the system integrates “movie” information over time is unclear. We characterized optic tectal neuronal receptive fields using sparse noise stimuli and reverse correlation analysis. Neurons recognized movies of ~200-600 ms durations with defined start stop stimuli. Movie from to responses were tuned sensory experience though a hierarchical algorithm. encoded families sequences following trigonometric functions. Spike sequence flow suggest that repetitive circuit motifs underlie movie detection. Principles frog topographic retinotectal plasticity cortical simple cells employed in machine learning networks for static recognition, suggesting discoveries principles encoding brain, such as duration encoded, may benefit recognition technology. built trained network mimicked neural encoders. The network, named MovieNet, outperformed current classifying natural scenes, while reducing data size steps complete classification task. This study reveals brain demonstrates brain-based processing enable efficient learning.

Language: Английский

Citations

0