International Journal of Neural Systems, Journal Year: 2025, Volume and Issue: unknown
Published: March 7, 2025
In this paper, we propose a spiking neural network model with Hebbian connectivity for implementing energy-efficient associative memory, whose activity is determined by input stimuli. The consists of three interacting layers Hodgkin–Huxley–Mainen neurons excitatory and inhibitory synaptic connections. Information patterns are stored in memory using symmetric matrix can be retrieved response to specific stimulus pattern. Binary images encoded in-phase anti-phase oscillations relative global clock signal. Utilizing the phase-locking effect allows cluster synchronization (both on output layers). Interneurons intermediate layer filter signal propagation pathways depending context layer, effectively engaging only portion connections within recognition. stability oscillation phase investigated both modes when recognizing direct inverse images. This context-dependent opens promising avenues development analog hardware circuits neurocomputing applications, potentially leading breakthroughs artificial intelligence cognitive computing.
Language: Английский
Citations
0Gravity-constrained simultaneous localization and mapping for suppressing map warping in complex large-scale environments
Integrated Computer-Aided Engineering, Journal Year: 2025, Volume and Issue: unknown
Published: April 13, 2025
Simultaneous Localization and Mapping (SLAM) serves as a foundational technology for autonomous systems operating within large-scale, complex environments. Traditional SLAM methodologies, however, are prone to altitude-axis distortions resulting from cumulative errors. To mitigate these issues, Gravity-Constrained (GC-SLAM) is introduced novel computational method that integrates gravity constraints incremental optimisation enhance mapping accuracy efficiency. GC-SLAM incorporates constraint handling actor the global algorithm, effectively reducing vertical-axis errors caused by accumulated drift during mapping. Furthermore, an strategy employed manage complexity associated with increasing map size. Performance evaluations of conducted on KITTI dataset large-scale environments, comparing its effectiveness against state-of-the-art SLAM-based algorithms, including FAST-LIO2, LIO-SAM (Lidar Inertial Odometry SLAM), Lego-LOAM (Lightweight Ground-optimised Lidar Mapping), A-LOAM (Advanced Mapping). Experimental results demonstrate suppresses distortions, significantly enhances localisation accuracy, outperforms competing methods.
Language: Английский
Citations
0Understanding joint range of motion development in robotic learning
Integrated Computer-Aided Engineering, Journal Year: 2025, Volume and Issue: unknown
Published: April 23, 2025
Joint Range of Motion (JROM) development has been shown to facilitate learning motor control in human beings. This developmental strategy applied robotics improve performance with different outcomes: sometimes it is favourable, others irrelevant, and others, even detrimental. The reasons that underpin this variability the results are still not well understood. In paper, we seek better understand principles underlying application JROM based morphological make its use more straightforward. To end, empirical studies were carried out over two representative cases: quadruped bipedal robot morphologies walk. Different parameters (morphological configuration, strategy, etc.) have evaluated elucidate their effects learning. show there significant connections between reduction space induced by way exploration exploitation solution algorithm, achieved. Through these connections, identified a set conditions must be satisfied for effective as tool improvement.
Language: Английский
Citations
0A solution to the only one object problem with dissolution rules
Journal of Membrane Computing, Journal Year: 2024, Volume and Issue: 6(2), P. 101 - 108
Published: May 1, 2024
Abstract
In
the
framework
of
membrane
computing,
(non-)uniform
families
recognizer
systems
are
usually
defined
to
solve
abstract
decision
problems.
this
sense,
use
finite
resources
for
each
member
family
makes
difference
with
respect
Turing
machines
solving
these
While
keeping
nature
systems,
it
is
interesting
know
which
type
problems
can
be
solved
by
means
a
single
system.
For
purpose,
complexity
class
$$\textbf{PMC}^{1p}_{\mathcal
{R}}$$
Language: Английский
Citations
2Unraveling the Differential Efficiency of Dorsal and Ventral Pathways in Visual Semantic Decoding
International Journal of Neural Systems, Journal Year: 2024, Volume and Issue: unknown
Published: Dec. 17, 2024
Visual semantic decoding aims to extract perceived information from the visual responses of human brain and convert it into interpretable labels. Although significant progress has been made in across individual cortices, studies on ventral dorsal cortical pathways remain limited. This study proposed a graph neural network (GNN)-based model natural scene dataset (NSD) investigate differences between process various parts speech, including verbs, nouns, adjectives. Our results indicate that accuracies for verbs nouns with motion attributes were significantly higher pathway as compared those pathway. Comparative analyses reveal outperformed terms performance attributes, evidence showing this superiority largely stemmed higher-level cortices rather than lower-level ones. Furthermore, these two appear converge their heightened sensitivity toward content related actions. These findings unique mechanisms through which segregate when processing stimuli different categories.
Language: Английский
Citations
0More Applications of Spiking Neural P Systems
Published: Jan. 1, 2024
Language: Английский
Citations
0