AUDEERING GMBH

Their core work

Audeering is a German AI company specializing in speech and audio analysis, with deep expertise in emotion recognition from voice, pathological speech processing, and multimodal signal interpretation. They develop technology that extracts meaning from how people speak — detecting emotions, health conditions, and cognitive states from audio signals. Their work spans healthcare applications (assessing speech disorders, supporting wellbeing interventions) and smart environment solutions (multimodal analytics for cities and workplaces), positioning them as a speech AI provider for both clinical and industrial use cases.

What they specialise in

How they've shifted over time

Audeering's early H2020 work (2015–2018) was tightly focused on voice and speech: emotion detection from voice (VocEmoApI), pathological speech processing (TAPAS), and wellbeing assessment (ECoWeB). From 2019 onward, they broadened significantly into multimodal AI — moving beyond audio-only into sign language translation (EASIER), smart city analytics (MARVEL), and distributed computing architectures. The shift shows a company expanding from a niche speech-AI specialist into a broader multimodal intelligence provider, applying their core audio expertise to new domains like accessibility and urban sensing.

Audeering is moving from pure speech/voice analysis toward multimodal AI systems that combine audio with visual and contextual signals — expect them to pursue projects in accessible communication, smart environments, and health monitoring.

How they like to work

Audeering operates primarily as a specialist contributor within larger consortia — they coordinated only one project (VocEmoApI, a small ERC Proof-of-Concept) and joined five others as participant or partner. With 74 unique consortium partners across 18 countries, they are well-networked and comfortable working in large European research teams. This pattern suggests they bring focused technical capability (speech/audio AI) into multidisciplinary projects rather than leading broad initiatives themselves.

Audeering has built a broad European network of 74 unique partners across 18 countries through just 6 projects, indicating they consistently join large, diverse consortia rather than working with a tight circle of repeat collaborators.

What sets them apart

Audeering occupies a rare niche at the intersection of speech AI and health — few SMEs combine deep audio emotion recognition expertise with clinical speech assessment capabilities. Their progression from voice emotion detection to sign language translation and multimodal city analytics shows they can adapt core audio intelligence to very different application domains. For consortium builders, they offer a commercially-minded SME (not a university lab) that can deliver production-grade speech and audio AI components within research projects.

Highlights from their portfolio

• VocEmoApI
  Their only coordinated project — an ERC Proof-of-Concept for voice emotion detection, representing the foundational technology behind most of their subsequent work.
• EASIER
  Marks their expansion beyond audio into visual-linguistic AI, applying neural machine translation and deep learning to automatic sign language translation.
• MARVEL
  Largest single EC contribution (EUR 329,850) and represents their move into smart city multimodal analytics with edge-fog-cloud architectures.