Predict
mood
better than any other AI

Our API platform offers science-backed expression and language models, built to amplify human well-being

Our models

Understand users' preferences by integrating language and expression

Capture nuances in expression—subtle facial movements of love or admiration, laughter tinged with awkwardness, sighs of relief—and build custom expression-language models

Expression Measurement
Speech Prosody
Model

Voice

Discover over 25 patterns of tune, rhythm, and timbre that imbue everyday speech with complex, blended meanings

Vocal Expression
Model

Voice

Differentiate 28 kinds of vocal expression recognized as conveying distinct meanings, and the many ways they are blended together

Facial Expression
Model

Face + Body

Differentiate 37 kinds of facial movement that are recognized as conveying distinct meanings, and the many ways they are blended together

Expressive Language
Model

Voice

Measure 53 emotions reliably expressed by the subtleties of language

Vocal Call Types
Model

Voice

Explore vocal utterances by inferring probabilities of 67 descriptors, like 'laugh', 'sigh', 'shriek', 'oh', 'ahh', 'mhm', and more

FACS 2.0
Model

Face + Body

An improved, automated facial action coding system (FACS): measure 26 facial action units (AUs) and 29 other features with even less bias than traditional FACS

Dynamic Reaction
Model

Face + Body

Measure dynamic patterns in facial expression over time that are correlated with over 20 distinct reported emotions

Toxicity
Model

Voice

Measure the toxicity of a sentence and the likelihood that it includes hate speech, an insult, an obscenity or a threat

Sentiment
Model

Voice

Measure the distribution of possible sentiments expressed in a sentence, negative to positive, neutral or ambiguous

Example Custom Models
Parkinson's Screening
Audio

Health & Wellness

Predicts the risk of Parkinson's from the voice with 32x fewer errors than a language model in data from a scientific study in Italy

Toxic Speech
Audio

Moderation

Identifies toxic speech by online gamers with a 36% lower error rate than a language-only model

Call Quality
Audio

Customer Service

Uses the voice to classify the quality of customer support calls with 2.8x fewer errors than language alone in data from Lawyer.com

Depressed Mood
Video

Health & Wellness

Infers the mood of vloggers with a 2.4x lower error rate by incorporating expression compared to language alone

Driver Drowsiness
Video

Safety

Classifies whether a driver appears to be alert or drowsy using publicly available video clips of real drivers and actors

Attentive vs. Distracted
Video

Training & Edtech

Predicts whether someone on a Zoom call appears to be paying attention or distracted, trained on publicly available recordings

Interaction Toolkit

A unified API for human understanding

Transcribe and understand language

Detect, transcribe, diarize, & understand speech in any language

Measure vocal and facial expression

Capture hundreds of expression dimensions per second in audio or video

Build custom models

Combine language and expression to understand your users’ preferences and needs

Grow with our model ecosystem

Predict user satisfaction & frustration, health & well-being, comprehension & confusion, & much more

Deploy and share models

Deploy your model instantly through our API to build it into your application or share it

Bootstrap with no-code tools

Analyze data, build & deploy models, and visualize results before writing any code

The new science of expression

Our scientific discoveries pave the way for AI to learn to make people happy

See all publications

No one can read minds, so we start by using statistics to ask what expressions mean to the people making them

Our data-driven science represents real emotional behavior with 3x more precision than traditional approaches

We use scientific control and globally diverse data to remove biases that are entrenched in most AI models

    1 of 3

The new science of human expression

Nature Human BehaviorScience AdvancesNatureTrends In Cognitive SciencesPNAS

The key to building beneficial AGI

Our mission is to ensure that artificial intelligence is optimized to serve human goals and emotional well-being

Our Blog

Latest blog posts

Our latest product updates, news, and tutorials

By using this website, you agree to our use of cookies.

Hume