|
Kategorie: Diplomové, bakalářské práce |
Tento dokument chci!
Práce popisuje základy principu funkčnosti neuronů a vytvoření umělých neuronových sítí. Je zde důkladně popsána struktura a funkce neuronů a ukázán nejpoužívanější algoritmus pro učení neuronů. Základy fuzzy logiky, včetně jejich výhod a nevýhod, jsou rovněž prezentovány. Detailněji je popsán algoritmus zpětného šíření chyb a adaptivní neuro-fuzzy inferenční systém. Tyto techniky poskytují efektivní způsoby učení neuronových sítí.
THE SPEECH SIGNAL
Speech/voice recognition difficult task performed computer
system [12].28
6.
A speaker can control and categorize sentence and make declarative,
interrogative imperative based the speaker’s purpose. Nonlinguistic information concerns idiosyncratic factors and emotional
states (such anger, sadness and delight) the speaker. Linguistic information can defined “symbolic information that is
represented set discrete symbols and rules for their combination“ [14].
Paralinguistic information defined “information that not inferable from a
written counterpart but deliberately added the speaker modify supplement
linguistic information” [14] and can have both discrete and continuous characteristics. Idiosyncratic factors which affect the
characteristics speech are age, gender, individual morphological characteristics,
health condition and possible physical handicaps. Pitch is
responsible for the tone the sound. Generally, the speaker
cannot control these factors, although possible for speaker imitate some
characteristics these factors actors [14].
The three deciding factors when talking about human-like perception of
speech are loudness, pitch and quality. The greater the amplitude is, the louder the sound appears. Each word sentence has specific
meaning and function and can divided into smaller segments: syllables and
phonemes. The quality sound perceptual correlate its spectral
content related the fundamental frequency the vocal vibration the speaker
organ [13]. Higher pitch issues higher tone and against,
lower pitches lower tone. Loudness represents the energy (intensity) of
the sound.
Speech sequence waves which are transmitted through medium and
are characterized some features, including characteristic frequencies and
corresponding intensities [13]. Although wide range commercial products were launched the last
decade, absolute solution has not been found out yet, and many research areas
have still remained opened the field.
.
The primary objective human speech communication transfer linguistic
information.
Speech communication crucial channel for conveying various kinds of
information that can divided into three categories terms its content: linguistic,
paralinguistic and nonlinguistic.
Besides linguistic and paralinguistic information, speech also contains nonlinguistic
information. The phoneme the smallest segment sound. The vibrations sound waves are perceived by
eardrums the inner ear, and these oscillations are forwarded specific part of
brain for further processing. The speech due the
effects paralinguistic information changing among neutral, admirable,
suspicious and disappointed states. An
important difference between linguistic and non-linguistic information that linguistic
information can controlled the speaker
