JPH05282273A - Device to discriminate event from behavior of organism - Google Patents

Device to discriminate event from behavior of organism

Info

Publication number
JPH05282273A
JPH05282273A JP10391692A JP10391692A JPH05282273A JP H05282273 A JPH05282273 A JP H05282273A JP 10391692 A JP10391692 A JP 10391692A JP 10391692 A JP10391692 A JP 10391692A JP H05282273 A JPH05282273 A JP H05282273A
Authority
JP
Japan
Prior art keywords
behavior
event
neural network
organism
dog
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Pending
Application number
JP10391692A
Other languages
Japanese (ja)
Inventor
Yoshihiro Matsumoto
善博 松本
Hidenobu Komatsu
英伸 小松
Kazuhiko Aoki
一彦 青木
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Nuclear Fuel Industries Ltd
Original Assignee
Nuclear Fuel Industries Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Nuclear Fuel Industries Ltd filed Critical Nuclear Fuel Industries Ltd
Priority to JP10391692A priority Critical patent/JPH05282273A/en
Publication of JPH05282273A publication Critical patent/JPH05282273A/en
Pending legal-status Critical Current

Links

Landscapes

  • Geophysics And Detection Of Objects (AREA)
  • Emergency Alarm Devices (AREA)

Abstract

PURPOSE:To provide a device capable of discriminating correctly an event being occurring from the behavior of an organism, and capable of serving communication with the organism other than a human being or the foreknowledge of the natural disaster, etc. CONSTITUTION:The behavior of a dog 4 (organism) is turned into a parameter by an input part 1 (barking by frequency analysis, movement of tail or ear, line of sight, time zone, etc.), and is inputted to a neural network system 2. In this neural network system 2, learning data (group of behavior and event) is given beforehand, and the weight factor of each node is adjusted beforehand, and the event being occurring is discriminated in accordance with an input signal. A discriminated result (fawning on person, presence of a suspicious person, etc.) is outputted from an output part 3, and thus, information the dog 4 wants to transmit can be known exactly. This device can be served to the foreknowledge of the natural disaster, etc., by applying it to the organism other than the dog.

Description

【発明の詳細な説明】Detailed Description of the Invention

【0001】[0001]

【産業上の利用分野】人間と共存する生物の挙動を分析
することにより、環境の変化等生じている事象を判別す
るための装置に関するものである。
[Field of Industrial Application] The present invention relates to a device for discriminating an event such as a change in environment by analyzing the behavior of a living organism coexisting with a human.

【0002】[0002]

【従来の技術】人間がペットとして犬や猫等の生物を飼
うことは一般的に行なわれており、人間がペットに特別
の愛着を感じたり、ペットが飼い主になつくなどのこと
がある。長年ペットを飼っている場合には、ペットの鳴
き声や尾の振り方などから、飼い主がペットのごく基本
的な喜怒哀楽(例えば危険を感じておびえている)や生
理的な欲求(例えば空腹でえさが欲しい)を理解するこ
とはある程度は可能である。
2. Description of the Related Art It is common for human beings to keep living things such as dogs and cats as pets, and there are cases in which humans feel a special attachment to pets, and pets become owners of pets. If you have a pet for many years, the owner's crying voice and how you swing your tail will affect the owner's very basic emotions (for example, feeling scared) and physiological needs (for example, hunger). It is possible to understand to some extent.

【0003】又、人間以外の生物は、特定の分野では人
間より優れた感覚を有していることが多く、例えば犬は
人間より格段に優れた嗅覚を有している。この他、地震
や山火事などにおいて動物が逸早く危険を察知したり、
植物が環境汚染に対して敏感に反応するなどのこともあ
る。
Living organisms other than humans often have sensations superior to humans in specific fields, and dogs, for example, have a significantly better sense of smell than humans. In addition, animals can quickly detect danger in earthquakes and forest fires,
There are also cases where plants react sensitively to environmental pollution.

【0004】[0004]

【発明が解決しようとする課題】しかし、従来において
は、ペットを飼う場合であっても、基本的に人間の欲求
を満たすことを目的としており、一方的な人間の感情の
押しつけである場合が多い。上述のように、個々の場合
においてペットの挙動から飼い主が生じている事象を推
測できる場合もあるが、飼い主側の勝手な思いこみであ
ることも多く、具体的なコミュニケーション手段は皆無
である。又、人間以外の生物が災害や環境汚染を逸早く
察知したとしても、従来においては人間への情報伝達が
行なわれておらず、被害を回避するために役立てること
ができなかった。
However, in the past, even in the case of keeping a pet, the purpose is basically to satisfy the human desire, and in some cases it is a one-sided imposition of human emotions. Many. As described above, in some cases, it may be possible to infer the event that the owner is taking from the behavior of the pet in each case, but this is often a self-intention of the owner, and there is no specific communication means. Even if organisms other than human beings quickly detect disasters and environmental pollution, information has not been transmitted to humans in the past, and it has not been possible to help prevent damage.

【0005】本発明は、上記の点に鑑みてなされたもの
であり、生物の挙動から生じている事象を正確に判別で
き、人間以外の生物とのコミュニケーションや天変地異
の予知等に役立てることのできる装置を提供することを
目的とするものである。
The present invention has been made in view of the above points, and is capable of accurately discriminating an event caused by the behavior of a living thing, and being useful for communication with living things other than human beings, prediction of natural disasters, and the like. The purpose is to provide a device that can.

【0006】[0006]

【課題を解決するための手段】本発明の装置は、生物の
挙動を分析することにより、生じている事象を判別する
装置であって、上記の課題を達成するために、前記生物
の挙動をパラメータ化した入力信号に基づいて前記事象
を判別する判別手段として、所定の応答関数f(X)を
もつ複数の処理ユニットが所定の重み係数Wi で結合さ
れ、かつ前記重み係数は与えられた学習データによって
最適化がなされるニューラルネットワークシステムを備
えたものである。
The device of the present invention is a device for discriminating a phenomenon occurring by analyzing the behavior of a living thing, and in order to achieve the above-mentioned object, the behavior of the living thing is As a discriminating means for discriminating the event based on the parameterized input signal, a plurality of processing units having a predetermined response function f (X) are combined with a predetermined weighting factor W i , and the weighting factor is given. It is equipped with a neural network system optimized by learning data.

【0007】[0007]

【作用】本発明においては上記のようにニューラルネッ
トワークシステムを用いて生物の挙動を分析して事象を
判別するが、このニューラルネットワークについて以下
に説明する。ニューラルネットワークとは、人間の神経
細胞を模擬するモデルであり、神経細胞(ニューロン)
に相当する人工の処理ユニットが所定の重み係数で結合
されている。ここでは、簡単のため、層構造の前進型ニ
ューラルネットワークを例にとり図2を参照して説明す
る。又、それぞれの処理ユニットはネットワークの交点
(ノード)にあたり、一般にノードと称されるので以下
においては処理ユニットをノードと表記する。
In the present invention, the behavior of a living being is analyzed by using the neural network system as described above to discriminate an event. This neural network will be described below. A neural network is a model that simulates human nerve cells.
An artificial processing unit corresponding to is connected with a predetermined weighting factor. Here, for the sake of simplicity, a forward neural network having a layered structure will be described as an example with reference to FIG. Further, each processing unit corresponds to an intersection (node) of the network and is generally called a node. Therefore, in the following, the processing unit is referred to as a node.

【0008】図2(A)において、ニューラルネットワ
ークは入力層9、中間層10、出力層11からなり、各
層は複数のノード8(u1 …u9 )から構成されてい
る。各層のノード8間はそれぞれ重み係数Wxx(後述)
で結合されている。このようなニューラルネットワーク
において、入力層9のノード8の数はパラメータ化され
た入力信号の数に応じて、出力層11のノード8の数は
出力信号の数(出力すべき項目の数)に応じて決められ
る。又、中間層については、層数、ノードの数ともに可
変である。
In FIG. 2 (A), the neural network comprises an input layer 9, an intermediate layer 10 and an output layer 11, each layer being composed of a plurality of nodes 8 (u 1 ... U 9 ). Between the nodes 8 of each layer, the weighting factor W xx (described later)
Are joined by. In such a neural network, the number of nodes 8 in the input layer 9 depends on the number of parameterized input signals, and the number of nodes 8 in the output layer 11 corresponds to the number of output signals (the number of items to be output). It is decided according to. Further, regarding the intermediate layer, both the number of layers and the number of nodes are variable.

【0009】次に、各ノード内部の処理について図2
(B)を参照して説明する。1つ前の層からの入力信号
(I1 …In )に対し、式1に示されるように、1つ前
の層に対する重み係数(W1 …Wn )がそれぞれ乗算さ
れ、更に重みづけされた入力値の総和が求められた後、
総和から閾値Hi が減算されてXが求められる。 X=ΣWi ・Ii −Hi …(1)
Next, the processing inside each node is shown in FIG.
This will be described with reference to (B). The input signals (I 1 ... I n ) from the previous layer are multiplied by the weighting factors (W 1 ... W n ) for the previous layer, respectively, as shown in Expression 1, and further weighted. After the sum of the input values
The threshold value H i is subtracted from the total sum to obtain X. X = ΣW i · I i −H i (1)

【0010】次いで、このXは、非線形関数であるシグ
モイド関数(式2参照)に代入されてf(X)が求めら
れる。 f(X)=(1+tan-1 (X/u0 ))/2 …(2) なお、ノード8の応答関数は、必ずしも上記のシグモイ
ド関数に限らず、他の関数であっても良いものである。
Next, this X is substituted into a sigmoid function (see equation 2) which is a non-linear function to obtain f (X). f (X) = (1 + tan −1 (X / u 0 )) / 2 (2) The response function of the node 8 is not limited to the sigmoid function described above, and may be another function. is there.

【0011】さて、ニューラルネットワークシステムを
使用するに際しては、まず、学習を行なう。ここで学習
とは、情報処理システムの目的に合うようにノード間の
重み(W1 …Wn )及び閾値Hj を調節することであ
る。学習の具体的な方法としては、入力(質問)と出力
(答)が既知である1組のデータを多数用意し、入力−
出力がすべて一致するようにノード間の重み係数(Wi
…Wn )及び閾値(Hj)を決定すれば良い。
When using the neural network system, learning is first performed. Here, learning means adjusting the weights (W 1 ... W n ) between the nodes and the threshold value H j so as to meet the purpose of the information processing system. As a concrete learning method, a large number of sets of data whose inputs (questions) and outputs (answers) are known are prepared and input-
The weighting factors (W i
... W n ) and the threshold value (H j ) may be determined.

【0012】一旦学習が終了すると、ニューラルネット
ワークシステムは「知能」をもつことになり、未知の入
力(質問)に対して各ノードの演算結果として出力
(答)がなされる。このように、ニューラルネットワー
クシステムでは、各ノードで並列処理が行なわれ、入力
から出力までの過程が簡単な演算で行なわれるので、処
理速度が非常に早い。又、一旦学習が終了しているもの
について、更に学習を行なうことができ(これにより重
み係数及び閾値がより最適な値に調節される)、順次評
価能力を向上させることができる。
Once the learning is completed, the neural network system has "intelligence", and an output (answer) is made as an operation result of each node with respect to an unknown input (question). As described above, in the neural network system, parallel processing is performed in each node, and the process from input to output is performed by a simple calculation, so that the processing speed is very fast. Further, it is possible to carry out further learning on the objects for which learning has been completed once (the weighting coefficient and the threshold value are adjusted to more optimum values), and the evaluation ability can be improved sequentially.

【0013】又、上記においては、層構造をなす前進型
ニューラルネットワークについて説明したが、より人間
の神経系に近い結合モデルである相互結合型ニューラル
ネットワーク(図3参照)も開発されており、これを使
用して測定信号の評価を行なうこともできる。図2の前
進型ニューラルネットワークでは、各層間の結合だけで
同じ層内のノード8は結合されていないが、相互結合型
ニューラルネットワーでは図3に示されるように、各ノ
ードが結合されている。
In the above description, the forward neural network having a layered structure has been described, but an interconnected neural network (see FIG. 3), which is a connected model closer to the human nervous system, has also been developed. Can also be used to evaluate the measured signal. In the forward neural network of FIG. 2, the nodes 8 in the same layer are not connected only by the connection between the layers, but in the mutual connection type neural network, the nodes are connected as shown in FIG. ..

【0014】[0014]

【実施例】本発明の実施例として、犬の挙動から事象を
判別する装置のブロック図を図1に示す。図において、
犬4の挙動は入力部1においてパラメータ化され、各パ
ラメータが数値化(ランク分け)される。この数値化
は、人間が別途に行なっても良いものであるが、本実施
例においては入力部1はビデオカメラ等の撮像手段やマ
イクロフォン、及び演算手段を備え、自動的に犬4の挙
動が検出されて数値化される。具体的には、撮像手段か
らの出力信号に基づいて犬4の尾や耳等の動きが数値化
され、マイクロフォンからの出力信号に基づいて4犬の
鳴き声の周波数解析がなされる。
1 is a block diagram of an apparatus for discriminating an event from the behavior of a dog as an embodiment of the present invention. In the figure,
The behavior of the dog 4 is parameterized in the input unit 1, and each parameter is digitized (ranked). This digitization may be performed by a person separately, but in the present embodiment, the input unit 1 is provided with an image pickup means such as a video camera, a microphone, and a calculation means, and the behavior of the dog 4 is automatically adjusted. It is detected and digitized. Specifically, the movements of the tail, ears, etc. of the dog 4 are digitized based on the output signal from the image pickup means, and the frequency analysis of the barks of the four dogs is performed based on the output signal from the microphone.

【0015】次に、犬4の挙動をパラメータ化した信号
は、ニューラルネットワークシステム2に入力され、こ
こで生じている事象の判別がなされる。本実施例におけ
るニューラルネットワークシステム2は、図2で説明し
た前進型ニューラルネットワークで構成されており、既
存のパソコンによって稼働される。
Next, the signal obtained by parameterizing the behavior of the dog 4 is input to the neural network system 2 and the event occurring here is discriminated. The neural network system 2 in this embodiment is composed of the forward neural network described with reference to FIG. 2, and is operated by an existing personal computer.

【0016】ニューラルネットワークシステム2による
判別結果は、出力部3から出力される。この出力部3
は、モニタ画面等に判別結果を表示すると共に、緊急の
事象が生じた場合(例えば不審者の侵入)のためにブザ
ー等の警報手段を備えている。
The result of discrimination by the neural network system 2 is output from the output unit 3. This output unit 3
Displays a discrimination result on a monitor screen and the like, and is provided with an alarm means such as a buzzer for an emergency event (for example, intrusion of a suspicious person).

【0017】さて、上述した装置を使用して犬4の挙動
から事象を判別するにあたって、ニューラルネットワー
クシステム2に学習を施し、前述したノード間の重み係
数(図2参照)を調節しておく必要がある。以下に、パ
ラメータの例と学習データの例を示す。 パラメータ: 1.周波数解析による鳴き声。 2.尾の挙動。 3.耳の挙動。 4.視線 5.時間帯。
Now, in order to discriminate an event from the behavior of the dog 4 using the above-mentioned device, it is necessary to perform learning on the neural network system 2 and adjust the above-mentioned weighting coefficient between nodes (see FIG. 2). There is. Below, the example of a parameter and the example of learning data are shown. Parameters: 1. Cry by frequency analysis. 2. Tail behavior. 3. Ear behavior. 4. Line of sight 5. Time zone.

【0018】 [0018]

【0019】上記のような学習データは、出力信号に相
当する事象を故意に起こし、その際の犬の挙動を観察す
ることによって行なうことができる。一旦学習を施した
後は、出力信号から犬の伝えたいことを的確に知ること
ができる。
The learning data as described above can be performed by intentionally causing an event corresponding to the output signal and observing the behavior of the dog at that time. After learning once, it is possible to know exactly what the dog wants to convey from the output signal.

【0020】なお、上記の例では、家庭で飼われている
犬を例にとって示したが、この他、有効な諸生物の挙動
をパラメータ化してニューラルネットワークシステムに
入力し、生じている事象を判別すれば、天変地異や災害
の予知に役立てることができる。例えば、夜行性の生物
からの情報による夜間の異常事態の発見や、嗅覚に優れ
た生物からの情報によるガス漏れ等の異常事態の早期発
見などが可能である。又、ねずみ等の挙動を分析するこ
とにより、地震や火山の爆発の予知に役立てることもで
きる。
In the above example, the dog kept at home is taken as an example, but in addition to this, the behaviors of various living organisms are parameterized and input to the neural network system to discriminate the occurring events. Then, it can be useful for predicting natural disasters and disasters. For example, it is possible to detect an abnormal situation at night based on information from a nocturnal creature, or to early detect an abnormal situation such as a gas leak based on information from a creature with a good sense of smell. Further, by analyzing the behavior of a mouse or the like, it can be useful for predicting an earthquake or an explosion of a volcano.

【0021】本発明では、対象とする生物や事象が変わ
っても装置自体は基本的に変更する必要がなく、ニュー
ラルネットワークに適当な学習を施しさえすれば対応す
ることができる。又、新たな学習を施すことが可能であ
るので、順次判別能力を高めることができると共に、生
物の加齢などによる挙動の変化にも容易に対応すること
ができる。
In the present invention, even if the target organism or event changes, the device itself does not basically need to be changed, and it can be dealt with as long as the neural network is appropriately learned. In addition, since new learning can be performed, it is possible to sequentially improve the discriminating ability and easily cope with a change in behavior due to aging of a living thing.

【0022】[0022]

【発明の効果】本発明では、ニューラルネットワークを
使用して生物の挙動を分析し、生じている事象を判別す
るので以下のような効果を奏する。 1.生物(ペット等)の喜怒哀楽を的確に知ることがで
き、より親密な関係を作ることができる。 2.嗅覚や聴覚など優れた能力を有する生物からの情報
により、不審者の侵入、ガス漏れの早期発見、あるいは
天災の予知などが可能となる。又、夜行性の生物の挙動
を分析すれば、夜間の異常事態の発見が可能となる。
According to the present invention, the behavior of a living being is analyzed by using a neural network and the occurring event is discriminated. Therefore, the following effects are obtained. 1. You can know the emotions of living things (pets, etc.) accurately, and you can make a closer relationship. 2. Information from organisms with excellent olfactory and auditory abilities enables intrusion of suspicious persons, early detection of gas leaks, and prediction of natural disasters. In addition, by analyzing the behavior of nocturnal creatures, it becomes possible to discover abnormal situations at night.

【図面の簡単な説明】[Brief description of drawings]

【図1】本発明実施例による装置の構成を示すブロック
図である。
FIG. 1 is a block diagram showing a configuration of an apparatus according to an embodiment of the present invention.

【図2】ニューラルネットワークを説明するための概念
図であり、(A)は前進型ニューラルネットワークの構
成を、(B)はノード内の処理を示す。
2A and 2B are conceptual diagrams for explaining a neural network. FIG. 2A shows a configuration of a forward neural network, and FIG. 2B shows processing in a node.

【図3】相互結合型ニューラルネットワークの構成を示
す概念図である。
FIG. 3 is a conceptual diagram showing a configuration of an interconnected neural network.

【符号の説明】[Explanation of symbols]

1…入力部、2…ニューラルネットワークシステム、3
…出力部、4…犬、8…ノード、9…入力層、10…中
間層、11…出力層。
1 ... Input unit, 2 ... Neural network system, 3
Output unit, 4 dog, 8 node, 9 input layer, 10 intermediate layer, 11 output layer.

Claims (1)

【特許請求の範囲】[Claims] 【請求項1】 生物の挙動を分析することにより、生じ
ている事象を判別する装置であって、 前記生物の挙動をパラメータ化した入力信号に基づいて
前記事象を判別する判別手段として、所定の応答関数f
(X)をもつ複数の処理ユニットが所定の重み係数Wi
で結合され、かつ前記重み係数は与えられた学習データ
によって最適化がなされるニューラルネットワークシス
テムを備えたことを特徴とする生物の挙動から事象を判
別する装置。
1. An apparatus for discriminating an event that has occurred by analyzing the behavior of a living thing, wherein as a discriminating means for discriminating the event based on an input signal that parameterizes the behavior of the living thing. Response function f
A plurality of processing units having (X) have predetermined weighting factors W i
A device for discriminating an event from the behavior of a living thing, comprising a neural network system which is combined with each other and the weighting factor is optimized by given learning data.
JP10391692A 1992-03-31 1992-03-31 Device to discriminate event from behavior of organism Pending JPH05282273A (en)

Priority Applications (1)

Application Number Priority Date Filing Date Title
JP10391692A JPH05282273A (en) 1992-03-31 1992-03-31 Device to discriminate event from behavior of organism

Applications Claiming Priority (1)

Application Number Priority Date Filing Date Title
JP10391692A JPH05282273A (en) 1992-03-31 1992-03-31 Device to discriminate event from behavior of organism

Publications (1)

Publication Number Publication Date
JPH05282273A true JPH05282273A (en) 1993-10-29

Family

ID=14366757

Family Applications (1)

Application Number Title Priority Date Filing Date
JP10391692A Pending JPH05282273A (en) 1992-03-31 1992-03-31 Device to discriminate event from behavior of organism

Country Status (1)

Country Link
JP (1) JPH05282273A (en)

Cited By (5)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0813186A2 (en) * 1996-06-14 1997-12-17 Masaomi Yamamoto Animal's intention translational method
WO2003015076A1 (en) * 2001-08-06 2003-02-20 Index Corporation Device and method for judging dog's feeling from cry vocal character analysis
EP1031228A4 (en) * 1998-08-26 2005-03-30 Avshalom Bar-Shalom Device and method for automatic identification of sound patterns made by animals
JP2006505852A (en) * 2002-11-08 2006-02-16 ディービーエス セキュリティ リミテッド Canine security system
US7800506B2 (en) 2005-03-17 2010-09-21 Eyal Zehavi Canine security system

Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH03156152A (en) * 1989-11-13 1991-07-04 Mitsubishi Heavy Ind Ltd Malfunction diagnosis device of engine

Patent Citations (1)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JPH03156152A (en) * 1989-11-13 1991-07-04 Mitsubishi Heavy Ind Ltd Malfunction diagnosis device of engine

Cited By (9)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
EP0813186A2 (en) * 1996-06-14 1997-12-17 Masaomi Yamamoto Animal's intention translational method
EP0813186A3 (en) * 1996-06-14 1998-10-07 Masaomi Yamamoto Animal's intention translational method
EP1031228A4 (en) * 1998-08-26 2005-03-30 Avshalom Bar-Shalom Device and method for automatic identification of sound patterns made by animals
WO2003015076A1 (en) * 2001-08-06 2003-02-20 Index Corporation Device and method for judging dog's feeling from cry vocal character analysis
US6761131B2 (en) 2001-08-06 2004-07-13 Index Corporation Apparatus for determining dog's emotions by vocal analysis of barking sounds and method for the same
JPWO2003015076A1 (en) * 2001-08-06 2004-12-02 株式会社インデックス Dog emotion discrimination device and method based on voice feature analysis
JP2006505852A (en) * 2002-11-08 2006-02-16 ディービーエス セキュリティ リミテッド Canine security system
JP4876219B2 (en) * 2002-11-08 2012-02-15 バイオ−センス テクノロジーズ (ビー.エス.ティー.)リミテッド Canine security system
US7800506B2 (en) 2005-03-17 2010-09-21 Eyal Zehavi Canine security system

Similar Documents

Publication Publication Date Title
National Academies of Sciences et al. Approaches to understanding the cumulative effects of stressors on marine mammals
Boughman et al. Greater spear-nosed bats discriminate group mates by vocalizations
Bartneck et al. To kill a mockingbird robot
US8626505B2 (en) Identifying and generating audio cohorts based on audio data input
US20070183604A1 (en) Response to anomalous acoustic environments
Kanev et al. Anomaly detection in wireless sensor network of the “smart home” system
Hanson et al. Age differences in the response of California ground squirrels (Spermophilus beecheyi) to conspecific alarm calls
CN106846729A (en) A kind of fall detection method and system based on convolutional neural networks
Grabarczyk et al. Anthropogenic noise affects male house wren response to but not detection of territorial intruders
Placer et al. A fuzzy-neural system for identification of species-specific alarm calls of Gunnison's prairie dogs
Rose et al. Visitor presence and a changing soundscape, alongside environmental parameters, can predict enclosure usage in captive flamingos
Schulz et al. Privacy at home: an inquiry into sensors and robots for the stay at home elderly
CN112106075A (en) Extending previously trained deep neural networks
Sloan et al. The more the scarier: adult Richardson’s ground squirrels (Spermophilus richardsonii) assess response urgency via the number of alarm signallers
JPH05282273A (en) Device to discriminate event from behavior of organism
Munoz et al. Optimal multisensory integration
Danial et al. On the realization of the recognition-primed decision model for artificial agents
Reichert et al. New dimensions for animal communication networks: space and time
Nehaniv The making of meaning in societies: Semiotic and information-theoretic background to the evolution of communication
JP2022538933A (en) Bat exhibition viewing system and method
RU2665264C2 (en) Intelligent system of intruder detection
Iorshase et al. A neural based experimental fire-outbreak detection system for urban centres
Reilly et al. Modeling ground soldier situational awareness for constructive simulation with rules
CN112435515B (en) Artificial intelligence education robot
Jarmon Can Your Dog Hear You? perceptions About Canine Hearing Loss and Noise Exposure

Legal Events

Date Code Title Description
A02 Decision of refusal

Free format text: JAPANESE INTERMEDIATE CODE: A02

Effective date: 19970218