REALEYES OU

Their core work

Realeyes is an Estonian technology SME specialising in automated emotion and sentiment recognition from facial expressions captured via standard camera inputs. Their core work sits at the intersection of computer vision, affective computing, and machine learning — building systems that can infer emotional and psychological states from visual and acoustic signals without controlled laboratory conditions. In the H2020 programme they contributed to projects on real-world sentiment estimation and multimodal biometric authentication using lip movement and speech analysis. Their commercial focus is on measuring human attention and emotional response, most notably applied to advertising effectiveness and media analytics.

What they specialise in

How they've shifted over time

Both H2020 projects began in 2015 and overlap almost entirely in time, so there is no meaningful chronological shift to identify from the project data alone. What the two projects together suggest is a consistent and focused interest in camera-based human signal analysis — one project pushing into sentiment and affective states, the other into security-relevant biometric authentication. Given the absence of keyword metadata and the limited project count, tracking evolution requires caution: we can say they entered H2020 already specialised rather than generalised, but any deeper trajectory is not visible from this data.

Their dual involvement in both affective computing (SEWA) and biometric security (SpeechXRays) suggests they are positioned to expand into identity verification, liveness detection, and access control — sectors where emotion AI techniques are increasingly applied.

How they like to work

Realeyes has participated exclusively as a consortium partner across both H2020 projects, never taking the coordinator role — a pattern consistent with a technology provider contributing a specific capability rather than driving research agendas. Their 16 unique partners across 6 countries over just 2 projects suggests moderately sized consortia with little partner overlap, indicating they bring specialist technology that different research groups want to include. Working with them likely means engaging a focused team that delivers a defined component rather than a broad research partner.

Realeyes has collaborated with 16 unique partners across 6 countries from just 2 projects — a relatively wide network for such limited participation, pointing to active involvement in international consortia. No specific geographic concentration is visible from the data, though their Estonian base and European project history suggest a primarily northern and western European network.

What sets them apart

Realeyes stands out as a commercial technology SME that brings production-grade emotion AI into academic research consortia — a rare combination where their software is tested at research scale while they gain scientific validation and access to labelled datasets. They are one of very few Estonian companies to operate in both ICT and Security H2020 pillars simultaneously, which speaks to the dual-use nature of facial and speech analysis technology. For a consortium builder, they offer a direct bridge between academic affective computing research and commercial deployment in media, advertising, or identity verification.

Highlights from their portfolio

• SEWA
  Their largest funded project (€492,625) and the closest to their commercial core — automatic sentiment estimation in real-world conditions is precisely the problem Realeyes was built to solve.
• SpeechXRays
  Demonstrates their reach into security and biometrics beyond sentiment analysis, combining acoustic and visual modalities for identity authentication — a strategically distinct capability from their primary market.