Articles tagged with Speech
A new study found that people's pronunciation of certain words changes after hearing a southern-accented talker, even if they didn't hear the actual sound. Participants who had never lived in the U.S. south still converged on a southern-like pronunciation, suggesting that expectations about accents shape our speech.
Researchers have developed a system that can capture the information in speech signals similar to how humans perceive speech. The method uses a matching pursuit algorithm and psychoacoustic principles to produce high-quality resynthesized speech signals with natural accents.
Studies using magnetic resonance imaging found a correlation between decreased blood flow in specific brain areas and tinnitus severity. Researchers at the University of Illinois Urbana-Champaign explored these links to better understand the causes of tinnitus.
Researchers at the University of Groningen have developed an AI system that can recognize indoor spaces with high accuracy by combining image and audio data. The system achieved a 70% accuracy rate in recognizing nine different types of indoor spaces, surpassing previous results.
A study published in EPJ Data Science explores the effectiveness of counter speech in curbing online hate and disinformation. The research reveals that organized counter speech appears to contribute to a more balanced public discourse by increasing the frequency of counter speech and decreasing that of hate speech.
A Dartmouth study found that people who respond faster to each other in conversations feel more connected. The research, published in the Proceedings of the National Academy of Sciences, used response time data to measure social connection between strangers and close friends.
UCSF researchers discovered a small cluster of neurons in the brain stem that regulates tempo and coordinates breathing with vocalization. This finding has implications for understanding speech pathologies, as altered wiring in this system may cause difficulties with speaking.
People with mild dementia struggle to understand speech in both quiet and acoustically challenging situations. Effective communication strategies include speaking clearly and slowly, reducing background noise, and providing contextual information.
A new theory suggests that stuttering is caused by anomalies in the brain's initiation circuit, which chooses a word to speak. This circuit is separate from the muscle-coordinating circuit, and its impairment leads to stuttering. Researchers believe this could lead to targeted treatments for stuttering with fewer side effects.
Researchers at the University of California, Davis and the University of Texas at Dallas found that listeners can reliably identify the gender of individual children as young as 5. They also discovered that identification of gender must take place jointly with the identification of age and likely physical size.
Researchers at Japan Advanced Institute of Science and Technology create method to estimate five-room acoustic parameters and speech transmission index using a short conversation, with potential applications in monitoring auditoriums during concerts and saving lives through smart speakers
New research published in eNeuro found that sharing happy stories increases feelings of closeness and synchronizes brain activity between the speaker and listener. Brain synchrony was linked to increased interpersonal closeness, particularly in regions involved in emotional processing and theory of mind.
Researchers will track oral health behaviors and caries risk in low-income urban young children over five years, collecting data on caregiver-reported habits, dental plaque scores, and social risk factors. The study aims to understand the impact of community-level factors on individual behaviors.
Researchers used MEG to measure brain activity while participants listened to simultaneous math equations and sentences, finding separate networks for each. The parietal lobe processed math and the temporal lobe processed language, with activity tracking every word and symbol.
A Tokyo University of Science researcher analyzed 8,000 infant names and found that common boys' name "大翔" has at least 18 readings, while girls' name "結愛" has at least 14. The study shows that parents use various readings, including non-existent ones, and abbreviate common readings.
Researchers analyzed spoken and sung verbal snippets with corresponding drum and song excerpts, finding high acoustic correlation between talking drums and Yorùbá language. The study also explores the connection between music and speech, shedding light on the relationship between auditory communication and language evolution.
The study analyzed two million parliamentary speeches in the UK and Ireland, finding that politicians use emotive language more in high-profile debates like PMQs to attract a wider audience. In low-profile debates, emotional rhetoric is less pronounced as politicians mostly address colleagues.
Researchers found that intermediate-sized respiratory droplets emitted while speaking can carry significant amounts of virus and remain suspended in air for minutes. These 'speech droplets' pose a considerable risk of SARS-CoV-2 transmission, particularly in enclosed spaces without proper ventilation.
Researchers have developed a deep-learning model that can track changes in speech coordination features to detect depression. The app aims to provide early intervention before severe depression sets in, potentially saving lives.
Researchers at Kennesaw State University investigated how various types of face coverings affect students' understanding of their teachers. They found that some masks performed well acoustically, but not as effectively against virus transmission.
Research from Washington University in St. Louis found that electric fish pause before sharing important information to increase listener response. The study reveals an underlying mechanism for how pauses allow neurons to recover from stimulation, making the following message clearer for the listener.
Scientists have identified a specific region of the bird brain linked to stuttering patterns, using zebra finches as a model system. A simple pharmacological treatment can stimulate rapid firing in this area, inducing stuttering-like patterns that resemble human speech dysfunctions.
Researchers equipped a Pepper robot with self-directed dialogue to facilitate better human-robot communication. The AI model improved the robot's ability to solve dilemmas and outperformed international standards for collaborative robots. By
Researchers create an approach to flexibly interpret and respond to Soldier intent derived from spoken dialogue with autonomous systems. The technology enables bi-directional conversational interactions between Soldiers and autonomous systems, allowing for hands-free operation to improve situational awareness.
Researchers developed a wireless implant to rapidly change adult songbirds' songs, expanding the toolbox for neuroscientists to study brain behavior and vocalization. The device, about the size of a dime, can modulate specific neurons during birdsong, causing pitch changes.
Researchers at City St George's, University of London, found that consumers view activist brands critically if they lack moral sensitivity, vision, and integration. To avoid being labeled as 'woke-washers', brands must demonstrate these traits when challenging free speech.
A recent study found that bilingual infants have a strong preference for infant-directed speech in their native language, while monolinguals also respond more strongly. The research, involving hundreds of babies from around the world, suggests that language familiarity plays a key role in shaping infant language development.
A recent study found that bilingual babies have a similar interest in infant-directed speech as monolingual babies, and are developmentally on track. The study showed that exposure to one language does not hinder the learning of another, and that parents can confidently babble to their babies in one or two languages.
Researchers created a tool that uses light reflection analysis to identify deepfake photos, achieving 94% accuracy. The technique exploits differences in reflected light patterns between real and fake images, providing a promising solution to combat deepfakes.
Researchers at Carnegie Mellon University tested AI-powered hate speech detection software on over 680,000 chess comments and found that 82% of flagged comments did not contain hate speech. The study suggests that words like 'black' and 'white' are triggering phrases for the classifiers.
A team of researchers at Carnegie Mellon University found that people are more successful at identifying language associated with Republican speech than Democratic speech patterns. The study used machine learning to scan the Congressional Record and presidential debate corpora to isolate linguistic variation between the two parties.
Scientists identify a unique voice prosody pattern associated with honest speech, recognized across languages and influencing memory recall. The study, published in Nature Communications, reveals how brain automatically processes this characteristic sound to convey truth-value or certainty.
A study by Bocconi University argues that the US Supreme Court's 'free market of ideas' metaphor is misleading due to concentrated platform power. The authors suggest framing Internet regulation around this concept to address issues like content moderation and user protection.
A University of Miami study found that a soft food diet can reshape language by changing how people speak and pronounce sounds. The research analyzed thousands of languages and the speech patterns of 10 celebrities, including Freddie Mercury and Michael Phelps.
A new study suggests that a bedside EEG test can aid in prognosis for unresponsive brain injury patients by measuring their response to speech. The strength of these responses is correlated with the patient's outcome and can help clinicians make effective treatment decisions.
Researchers found that disposable surgical masks provide the best acoustic performance among tested masks, reducing high-frequency sound muffle and allowing speech to be heard clearly. Loosely woven cotton masks also perform well but may not block respiratory droplets as effectively.
Researchers found that acetylcholine, a neuromodulator, enhances neural discrimination of tones from noise stimuli, potentially contributing to processing important acoustic signals like speech. The study also identified novel connections between the modulatory centers and the auditory system.
Researchers found that listeners can accurately identify regional dialects and racial backgrounds from audio clips of English speakers. The study used 144 unique audio clips to categorize speakers into Midland, New York City, Southern U.S., Asian American, Black/African American, or white groups.
Researchers explored the impact of eye position on speech intelligibility within a cocktail party setting. They found that performance was best when eye fixation was on target and suffered when off target.
Researchers found that surgical and N95 masks can minimize negative effects on speech intelligibility, making them ideal options for instructors. By wearing these types of masks, educators can help students understand speech more effectively while ensuring their own safety.
Despite excellent auditory capacities, dogs fail to attend to differences between words with single speech sound variations. This may be why dogs' learned vocabulary remains low throughout life. Brain activity studies show dogs quickly differentiate known from nonsense words but not those differing in just one sound.
Researchers have developed a machine learning model that analyzes speech and facial movements to diagnose strokes with high accuracy, potentially saving valuable time in diagnosing the condition. The tool was tested on over 80 patients and achieved 79% accuracy comparable to clinical diagnostics by emergency room doctors.
Researchers discovered that speech generates conical airflow that can travel up to 2 meters in 30 seconds, emphasizing the importance of time and distance in transmission. Saliva filaments form on the lips during consonant sounds, breaking into microscopic droplets that spread pathogens.
A study by Kansas State University analyzed nearly 2 million congressional speeches and found that politicians use simpler language and express more positive or negative sentiments. The analysis also showed a decrease in reading level and vocabulary over time.
A new study reveals that dog brains, like human brains, process speech in a hierarchical manner, with intonation processed at lower stages and word meanings at higher stages. The research used functional MRI to measure brain activity in awake dogs while they listened to praise words and unknown neutral words in different tones.
A study of 90 participants found that watching a magician perform a card trick resulted in unconscious decisions influenced by verbal and nonverbal primes. The results suggest that embedding priming within speech and gestures can shape decision-making processes.
A new USC study has created a hate speech classifier that is more context-sensitive, reducing errors that amplify racial bias. The algorithm considers context and specific features of hate speech to improve accuracy.
A recent study found that the human brain processes speech with exceptional temporal fidelity, unlike other sounds. The researchers used millisecond-scale brain recordings to decode and reconstruct spoken words, discovering a unique time-locked encoding mechanism for speech.
Research at Goethe University Frankfurt reveals that both hemispheres of the brain contribute to speech control, with the left hemisphere handling fast processes and the right hemisphere controlling slower processes. This discovery challenges the long-held assumption that the right hemisphere only analyzes spoken words.
Researchers found chimpanzees produce lip-smacks at a speech-like rhythm of 4.15 Hz, confirming the existence of a continuous path in the evolution of primate mouth signals with this rhythm. This discovery supports the theory that spoken language evolved from existing primate communication systems.
Researchers found that participants synchronized their arm movements with those of the vocalizer, even with subtle and hard-to-detect movements. The study suggests that musculoskeletal tensioning affects speech, allowing for communication cues like excitement to be perceived through gesture-induced acoustics.
Researchers at Penn State developed a public-speaking tutor on Amazon Alexa, which helps users recognize and modify negative thinking behaviors. The study found that the tutor eased pre-speech anxiety in participants.
Research suggests that asymptomatic individuals infected with coronavirus can produce aerosolized particles during normal speech, potentially transmitting the infection. The study highlights the importance of collaboration among aerosol scientists, virologists, and engineers to understand respiratory diseases like COVID-19.
A new study from University of California, Riverside, found that Donald Trump's racially inflammatory speech emboldened particular members of the public to express their prejudices. Exposure to such rhetoric made individuals feel more comfortable expressing their prejudice, while silence from other elites gave the impression that a nor...
A University of York study reveals that community question answering sites like Yahoo! Answers are plagued by hate speech and Islamophobia when discussing terrorism. The research found sensitive questions were more likely to be submitted anonymously and loaded with negative emotions.
Researchers used CT scans and 3D printing to recreate the vocal tract of a 3,000-year-old Egyptian mummy, producing a single sound that falls between 'bed' and 'bad'. The achievement has implications for understanding past voices and presentation in the present.
Researchers developed an AI system to analyze hundreds of thousands of comments on social media, identifying positive comments defending disenfranchised minorities. The system can reduce manual efforts to find these 'help speech' comments by up to 88%.
Researchers found that brain activity in healthy adults synchronized with sound waves produced by a silent woman speaking, indicating the brain can process auditory information from visual cues. This ability arises from visual cortex activity synchronizing with lip movement and signal transmission to other areas for sound synthesis.
A researcher analyzed the frequency and acoustics of belching while speaking in 'Rick and Morty' to uncover latent linguistic meaning. The study found that burps tend to rumble at a low 300 hertz, jitter 4% more than normal speech, and shimmer 15% more.
A Yale study reveals that hiring managers judge candidates' competence and fitness for a job based on socioeconomic position inferred from brief speech patterns. The study shows that these snap perceptions influence hiring decisions in ways that perpetuate inequality.