Patents
Literature
2 results about "Habituation" patented technology
Efficacy Topic
Property
Owner
Technical Advancement
Application Domain
Technology Topic
Technology Field Word
Patent Country/Region
Patent Type
Patent Status
Application Year
Inventor
Habituation is a form of non-associative learning in which an innate (non-reinforced) response to a stimulus decreases after repeated or prolonged presentations of that stimulus. Responses that habituate include those that involve the intact organism (e.g., full-body startle response) or those that involve only components of the organism (e.g., habituation of neurotransmitter release from in vitro Aplysia sensory neurons). The broad ubiquity of habituation across all biologic phyla has resulted in it being called "the simplest, most universal form of learning...as fundamental a characteristic of life as DNA." Functionally-speaking, by diminishing the response to an inconsequential stimulus, habituation is thought to free-up cognitive resources to other stimuli that are associated with biologically important events (i.e., punishment/reward). For example, organisms may habituate to repeated sudden loud noises when they learn these have no consequences. A progressive decline of a behavior in a habituation procedure may also reflect nonspecific effects such as fatigue, which must be ruled out when the interest is in habituation. Habituation is clinically relevant, as a number of neuropsychiatric conditions, including autism, schizophrenia, migraine, and Tourette's, show reductions in habituation to a variety of stimulus-types both simple (tone) and complex (faces).
Self-evolution intelligent bird repelling method based on bird behavior modal and reinforcement learning
The application discloses a self-evolution intelligent bird repelling method based on bird behavior modal and reinforcement learning, and belongs to the technical field of wild animal prevention and artificial intelligence. In view of the technical pain point that traditional sound wave bird repelling equipment is prone to causing birds to develop "auditory habituation", the application proposes that bird behavior characteristics and micro-meteorological data are collected in real time through a multi-modal sensor, and a generative adversarial network (GAN) is introduced to dynamically generate non-repetitive adversarial voice prints. On the repelling strategy, a dynamic game model based on a deep Q network (DQN) is constructed, the escape acceleration, escape speed and stay time of the birds are quantified as a repelling efficiency index, and the repelling efficiency index is taken as feedback input into a reward function R(s, a)=α·A escape +β·V escape ‑γ· T linger , and the driving strategy is self-optimized. Meanwhile, the application combines a micro-meteorological pre-compensation mechanism to correct sound wave attenuation, and realizes model iteration of multiple nodes through cloud edge cooperation and federated learning. The application breaks the limitation of traditional mechanical repelling, constructs a self-evolution closed-loop system of "perception-evaluation-generation-adversarial", effectively solves the problem of bird neural adaptation, and realizes long-acting, precise and intelligent bird repelling effect.
Owner:DALI UNIV
A memristor decision circuit that fuses emotional habituation with behavioral reinforcement
The application discloses a kind of fusion mood habituation and behavior reinforcement's memristor decision circuit, the circuit includes positive signal input path and negative signal input path, and each path includes input module, mood generation module, decision output module, threshold adjustment module and result feedback module.Input module judges external signal polarity, when matching, the output signal of input module is applied to mood generation module to generate mood signal, mood signal is input to decision output module to make decision, and result feedback module generates different feedback signals according to whether the correctness of decision result, and the resistance and threshold adjustment module of memristor are adjusted.Multiple decisions, two signal input paths form decision habit, habituation reacts to mood generation, and the mood after habit formation tends to be stable.The proposed memristor decision circuit simulates the process of emotion-driven behavior, habit formation, habit modulates emotion, and can be used to realize brain-like intelligent decision.
Owner:HUNAN NORMAL UNIVERSITY