AI Face Intelligence

Facial Recognition & Reverse Image Search

Upload any photo and let IRBIS AI scan the public web — matching faces across social networks, news sources, and online profiles to surface identity information in seconds.

AI-powered matching

Public sources only

ISO 27001 certified

Get Started Free API Docs

Face detected

97% confidence

3 matches found

Facial Recognition Scan

IRBIS AI Engine

Scanning…

Capabilities

What the facial recognition API returns

Submit a photo and receive a structured set of identity signals — matched profiles, source URLs, confidence scores, and contextual data from across the public web.

Face Detection & Landmark Mapping

AI identifies and maps facial landmarks from the uploaded image — establishing a geometric signature used for cross-referencing against public sources.

Public Web Matching

The face signature is compared against publicly available images across social networks, news sites, professional directories, and open web sources.

Confidence Scoring

Each match is returned with a confidence percentage — allowing your system to filter by match quality threshold and prioritize high-confidence results.

Profile & Source URLs

Matched results include direct source URLs, platform names, and page context — giving investigators and analysts actionable links, not just abstract scores.

Identity Cross-Referencing

Combine facial match results with IRBIS email, phone, or name lookup to build a complete, verified identity profile from a single image.

Privacy-First by Design

IRBIS only searches publicly available images. No private databases, no biometric storage. Every search is confidential and GDPR-compliant.

How it works

From photo to identity in four steps

Upload a photo

Submit any image via the IRBIS API — JPG, PNG, or URL. The AI detects and extracts the face automatically

Map facial landmarks

The AI engine maps key facial geometry points — eyes, nose, jawline — building a unique match signature

Scan public sources

The signature is compared against public web images, social profiles, news archives, and open directories

Receive match results

JSON response returns matching URLs, platforms, confidence scores, and page context for each result found

Who uses it

Built for teams that need to verify identity from an image

Facial recognition search is foundational for any workflow where you need to connect a photo to an identity — from KYC to investigations to fraud prevention.

KYC & Identity Verification

Verify that a selfie or document photo matches publicly available images of the claimed identity. Add a biometric signal layer to your onboarding without building your own face infrastructure.

Investigators & OSINT

Find all publicly available online presence of an individual using only a photo. Ideal for due diligence, background research, and open-source investigations.

Finance & Fraud Prevention

Detect identity fraud by cross-referencing submitted photos against public records. Flag cases where the uploaded face does not match the claimed identity's online presence.

Media & Press Verification

Verify the identity of individuals in press photos, social media content, or uploaded images — confirming whether the person matches their claimed public profile.

E-Commerce & Marketplace Trust

Verify that seller or buyer profile photos correspond to real, traceable individuals — reducing the risk of fake accounts and identity impersonation on your platform.

HR & Background Screening

Confirm applicant identities during remote hiring by cross-referencing submitted headshots against their claimed professional online presence on LinkedIn and other platforms.

API response fields

Every field returned by the face API

Structured JSON response — source URLs, confidence scores, platform context, and face metadata in one call.

Face detection status & landmark countAlways returned

Match confidence score (0–100%)Always returned

Source URL & platform nameAlways returned

Name & profile context (when found)When available

Image metadata & page titleWhen available

Total match count & job statusAlways returned

face-search-response.json

// POST /api/developer/face_webscan { "id": 7732901, "status": "completed", "face_detected": true, "landmarks": 68, "matches": [ { "url": "linkedin.com/in/j-anderson", "platform": "LinkedIn", "confidence": 0.97, "name": "J. Anderson" }, { "url": "twitter.com/janderson", "platform": "Twitter", "confidence": 0.91, "name": "J. Anderson" } ], "total_matches": 3 }

FAQ

Common questions

How does the facial recognition search work?

You submit a photo via the IRBIS API. The AI maps facial landmarks to create a unique match signature, then compares it against publicly available images across social networks, news sources, and the open web — returning matched URLs and confidence scores.

What sources does IRBIS search?

IRBIS searches publicly available images only — social media profiles, professional directories, news archives, and open web sources. No private databases or biometric repositories are accessed.

How accurate is the face matching?

The AI engine achieves up to 97% confidence on clear, front-facing images. Each result includes a confidence score so you can apply your own threshold filter. Accuracy varies with image quality, angle, and lighting.

Is facial search legal and GDPR-compliant?

IRBIS searches only publicly available images and does not store biometric data. ESPY is GDPR-compliant and ISO 27001 certified. Users are responsible for ensuring their use of the results complies with applicable laws in their jurisdiction.

Can I submit image URLs instead of uploads?

Yes. The IRBIS Face API accepts both base64-encoded image data and direct image URLs. For high-volume pipelines the async job model is recommended — submit the request, receive a job ID, and poll for results.

How long does a facial search take?

Face detection and landmark mapping complete in under a second. Full public web matching typically completes within 20–40 seconds via the async job model, depending on the breadth of results found.