Hume provides the AI toolkit to measure, understand, and improve how technology affects human emotion. Use our models to measure human expressive behavior in images, videos, audio, and text and build applications that better predict users’ intents, preferences, and outcomes. Coming soon: better LLMs, aligned with real signs of user satisfaction and well-being.

Aegis Ventures

Wisdom Ventures

Comcast Ventures

Northwell Holdings

Union Square Ventures

LG Technology Ventures

Evan Sharp

Hume AI

Our emotional expressions are the building blocks of empathy and social connection. But for a long time, scientists have sorted them into only a handful of categories, greatly underestimating their complexity. 

At Hume AI, we’re building a new science of expression that measures the full spectrum of nuanced expressive behaviors people form every day. We’ve run experiments around the world and discovered that humans worldwide share more than 25 dimensions of meaning in our facial and vocal expressions. 

We’re translating these insights into the most accurate machine learning models ever built for understanding human expression. And we’re helping developers around the world build technology that supports human emotional well-being.

Want to learn more? Watch the video below!

Hume's expression measurement models are derived from semantic space theory (SST), a revolutionary computational approach to mapping out emotional behaviors in a dimensional space pioneered by our CEO & Chief Scientist Alan Cowen.

Before SST, researchers relied on outdated classification methods like the Basic 6 emotions and Facial Action Coding System (FACS), which is limited by its reliance on the human eye's capacity to discern features, often leading to subjective interpretations. By introducing computational methods to emotion science, we’ve discovered that human emotion is high-dimensional, blended, and categorizable.

In our latest blog by Jeff Brooks, discover why textbooks are being rewritten around semantic space theory: https://lnkd.in/ejMmiwWD

Hume AI shows that our feelings are way more complicated than just being happy, sad, or scared. Using Semantic Space Theory (SST), we've learned that there are over 20 distinct emotional dimensions beyond the basic six categories. This data-driven approach offers a nuanced understanding and computational modeling of emotions across cultures.

https://lnkd.in/e4vGkDpT

"What we cherish" is an innovative project by Gehl that uses Hume's models to explore the emotional landscape of our cities. It shows how AI can translate soft, qualitative data into deep visual insights — and turn feelings into art.  

Gehl, a leading architecture and research consultancy that focuses on putting life at the center of urban design, asked people to record stories on one simple question: what do you love about your neighborhood?

But communication is about so much more than just what we say — it's how we say it. That's why Gehl used Hume's AI models to decode the emotional expressions in speech prosody. We need to interpret the rhythm, stress, and intonation in speech for a richer understanding of community stories. Learn more about our models, available through a public API: 
https://lnkd.in/gXAQGiWF

Imagine seeing the feelings of your community. That's what Gehl did next: they visualized our expression measurements, turning each emotional tone into a soft shape with a musical note. This creates a unique audiovisual symphony.

This project celebrates the emotional richness of our communities by helping us understand, visualise, and cherish our connections. Explore more:
https://lnkd.in/gfb7FsWf

Affectiva is a company that offers emotion recognition technology powered by artificial intelligence (AI). It provides data analytics solutions for various industries.

Affectiva

Retorio is an AI-powered video recruiting platform that utilizes machine learning to analyze and evaluate job candidates' soft skills and personality traits.

Retorio

Integrated Personalized Medicine is a company that leverages AI to provide personalized medical solutions and data-driven insights for healthcare.

Integrated Personalized Medicine

Amazon Rekognition is a deep learning-based image and video analysis service that offers facial recognition, object detection, and content moderation capabilities.

Amazon Rekognition

Cognitive Mill is a company that provides AI-powered solutions for natural language processing, text analysis, and sentiment analysis.

Cognitive Mill

Exadel CompreFace is an open-source facial recognition platform that offers face detection, recognition, and attributes analysis.

Exadel CompreFace

NaturalReader is a text-to-speech software that uses natural-sounding voices to convert written text into spoken words.

NaturalReader

Kits AI is an AI-powered platform that offers various tools and solutions for data analysis, visualization, and predictive modeling.

Kits AI

FakeYou is a platform that utilizes deep learning technology to generate synthetic speech from text, including the creation of deep fake audio content.

FakeYou - Deep Fake Text to Speech

Hume AI is an AI-powered data annotation platform that enables businesses to create high-quality training data for machine learning models.

news impact assignment basic	general	May 2, 2024 9:22 PM	a043ea55
public peers	general	May 2, 2024 9:22 PM	440893a3
news daily refresh	general
news processing daily refresh	general
news impact assignment daily	general
news processing basic	general	May 2, 2024 9:22 PM	f2dcee56
company summary	general
signals	general
crunchbase trigger	general
crunchbase profile	general
li profile	general	May 2, 2024 9:22 PM	37e7e1ab
translate crunchbase fundingrounds	general
peers	general
news basic	general	May 2, 2024 9:22 PM	fbf94817
bundesanzeiger	filings
merger acquisitions	company_analysis
financials download nd	financials
private financials	financials

AI:HUME

Hume AI