A Moving Target Detection Model Inspired by Spatio-Temporal Information Accumulation of Avian Tectal Neurons DOI Creative Commons
Shuman Huang, Xiaoke Niu, Zhizhong Wang

и другие.

Mathematics, Год журнала: 2023, Номер 11(5), С. 1169 - 1169

Опубликована: Фев. 27, 2023

Moving target detection in cluttered backgrounds is always considered a challenging problem for artificial visual systems, but it an innate instinct of many animal species, especially the avian. It has been reported that spatio-temporal information accumulation computation may contribute to high efficiency and sensitivity avian tectal neurons detecting moving targets. However, its functional roles are not clear. Here we established novel computational model The proposed mainly consists three layers: retina layer, superficial layers optic tectum, intermediate-deep tectum; last which motion would be enhanced by process. validity reliability this were tested on synthetic videos natural scenes. Compared EMD, without process accumulation, satisfactorily reproduces characteristics response. Furthermore, experimental results showed significant improvements over existing models (EMD, DSTMD, STMD plus) STNS RIST datasets. These findings do only understanding complicated processing avians, also further provide potential solution targets against environments.

Язык: Английский

A Moving Target Detection Model Inspired by Spatio-Temporal Information Accumulation of Avian Tectal Neurons DOI Creative Commons
Shuman Huang, Xiaoke Niu, Zhizhong Wang

и другие.

Mathematics, Год журнала: 2023, Номер 11(5), С. 1169 - 1169

Опубликована: Фев. 27, 2023

Moving target detection in cluttered backgrounds is always considered a challenging problem for artificial visual systems, but it an innate instinct of many animal species, especially the avian. It has been reported that spatio-temporal information accumulation computation may contribute to high efficiency and sensitivity avian tectal neurons detecting moving targets. However, its functional roles are not clear. Here we established novel computational model The proposed mainly consists three layers: retina layer, superficial layers optic tectum, intermediate-deep tectum; last which motion would be enhanced by process. validity reliability this were tested on synthetic videos natural scenes. Compared EMD, without process accumulation, satisfactorily reproduces characteristics response. Furthermore, experimental results showed significant improvements over existing models (EMD, DSTMD, STMD plus) STNS RIST datasets. These findings do only understanding complicated processing avians, also further provide potential solution targets against environments.

Язык: Английский

Процитировано

3