Lithuanian Speech Recognition Corpus (Mobile)

This dataset was recorded in a quiet office environment, with the involvement of 201 speakers, comprising 99 males and 102 females. All participants in the recording were carefully selected by professionals to ensure standardized pronunciation and clear enunciation. The recorded texts span various types of information, including news and everyday conversations.
Specifications:
ID:
King-ASR-707
Size:
258.8 hours
Language:
Lithuanian
Country
Lithuania
Sample rate & bit depth
16 kHz,16bit
Recording environment
Quiet (office)
Speaker
99 males and 102 females
Devices:
Mobile
Accuracy Rate
Sentence Accuracy Rate (SAR) 95%
Samples
Audio

People also searched for

Chinese Mandarin Speech Recognition Corpus for the Elderly and Children
This dataset is specifically tailored to capture the nuances of speech from the elderly and children, two demographic groups with distinct vocal characteristics. This dataset is recorded using desktop equipment to ensure high audio quality, and all recordings take place in a quiet environment to minimize background noise. The corpus includes read speech, which is beneficial for training speech recognition models on clear and deliberate pronunciations. The gender balance in the dataset ensures that the recognition system can accurately interpret both male and female voices. Furthermore, the speakers are drawn from the seven major Chinese dialect regions, providing a diverse and balanced distribution of accents and speech patterns. For the children's recordings, the dataset includes speech samples from interactive car control systems, children's audiobooks, children's video content, and music featuring children's songs and popular tunes from platforms like TikTok. The elderly recordings cover similar domains with a focus on applications and content that cater to their preferences, such as car control, map navigation, audiobooks with programs selected for an older audience, and music that includes selections favored by the elderly. This comprehensive approach ensures that the speech recognition system can effectively adapt to the unique speech traits of these age groups across various contexts and dialects.
Chinese Mandarin Speech Recognition Corpus
This dataset is a specialized collection of bilingual Chinese-English speech recordings, tailored to cater to the needs of speech recognition technology development. It is characterized by its unique blend of languages, high-fidelity audio quality, and the diversity of its contributors.
Chinese-English Mixed Speech Recognition Corpus (Desktop)
The dataset comprises speech samples that include a mixture of Mandarin Chinese and English, reflecting the linguistic diversity found in various global contexts. Each recording in the dataset is of superior quality, with clear and distinct sound that is essential for training robust speech recognition algorithms.
Chinese-English Mixed Speech Recognition Corpus
This dataset is designed to support the development and refinement of bilingual Chinese-English mixed speech recognition technologies. It contains a diverse set of speech samples recorded in various scenarios to train and test speech recognition systems. Recordings were primarily made using desktop devices to mimic everyday usage environments. All recordings are made using high-fidelity equipment to ensure clarity and improve the accuracy of speech recognition systems. Recordings were conducted in noise-free or low-noise environments to minimize the impact of background noise on speech recognition performance. Speakers are selected from the seven major Chinese dialect regions to achieve a balanced representation of regional accents.

Join our newsletter to stay updated

Thank you for signing up!

Stay informed and ahead with the latest updates, insights, and exclusive content delivered straight to your inbox.

By subscribing you agree to with our Privacy Policy and provide consent to receive updates from our company.