相互影响
发表于 2025-3-25 06:13:20
http://reply.papertrans.cn/88/8765/876449/876449_21.png
incredulity
发表于 2025-3-25 08:00:45
Long-Term Statistical Feature Extraction from Speech Signal and Its Application in Emotion Recognitway for capturing the long-term properties of this signal. Having a fixed-length representation for a variable-length pattern like speech which preserved the task-relevant information allows for using a wide range of powerful discriminative models which could not effectively handle the variability i
蛙鸣声
发表于 2025-3-25 13:24:54
Rhythm-Based Syllabic Stress Learning Without Labelled Data, generated from a multiscale model of rhythm perception. The model outputs a sequence of events, corresponding to the sequences of strong-weak syllables present in speech, based on which a stressed/unstressed decision is taken. We tested our approach on two languages, Catalan and Spanish, and we fou
A精确的
发表于 2025-3-25 19:47:13
http://reply.papertrans.cn/88/8765/876449/876449_24.png
Sedative
发表于 2025-3-25 22:46:55
0302-9743 e and Speech Processing,SLSP 2015, held in Budapest, Hungary, in November 2015..The 26 full papers presented together with twoinvited talks were carefully reviewed and selected from 71 submissions. Thepapers cover topics such as: anaphora and coreference resolution; authorshipidentification, plagiar
NIP
发表于 2025-3-26 02:29:11
http://reply.papertrans.cn/88/8765/876449/876449_26.png
Cholagogue
发表于 2025-3-26 07:28:54
Decoding Distributed Tree Structures,rees (DT) [.], and analyze the reverse problem: can the original structure be reconstructed given only its distributed representation? Our experiments show that this is indeed the case, DT can encode a great deal of information of the original tree, and this information is often enough to reconstruct the original object format.
Ornament
发表于 2025-3-26 10:39:05
Semantic Features for Dialogue Act Recognition, COALS semantic spaces. An interesting contribution is that all the features are created without any supervision. These approaches are evaluated on a Czech dialogue corpus. We experimentally show that all proposed approaches significantly improve the recognition accuracy.
chandel
发表于 2025-3-26 14:59:49
http://reply.papertrans.cn/88/8765/876449/876449_29.png
Nonflammable
发表于 2025-3-26 20:35:46
Discourse Particles in French: Prosodic Parameters Extraction and Analysis,e speech corpora. A preliminary test concerning the automatic detection of the word function is also carried out using prosodic parameters only, leading to an encouraging result of 70 % correct identification.