Offline Face Forensics for Mac and Windows.

Built for private investigators and face forensics teams working with case files, tips, open-source media, and device extractions. Extract faces from large image and video collections, group them into “people,” and search your dataset — entirely offline. No cloud, no telemetry, no accounts.
Offline No cloud No telemetry No accounts
Why DawaImg
“Case media is fragmented across devices, screenshots, social platforms, and video — and reviewing faces at scale is slow, inconsistent, and hard to document.”
DawaImg keeps review local and structured: import case media, extract faces, cluster into “people,” run similarity searches, and export results — while preserving provenance back to original files and frames.

⏱️ 60-Second Overview

What it is
  • Desktop face forensics and media organization tool
  • Runs entirely on your machine
  • Designed for large image, video, and document collections
What you do
  1. Import folders, images, video, or PDFs
  2. Extract and cluster faces into “people”
  3. Search and compare faces locally
  4. Export faces with provenance back to source frames/files
  5. Generate case-ready review outputs
Evidence model
  • Non-destructive processing
  • Immutable source records
  • Frame/page-level provenance tracking
  • Parent-child artifact linkage
What it is not
  • No cloud processing or uploads
  • No preloaded databases
In short: you bring the case media, DawaImg processes it locally using a structured, non-destructive record model, and every extracted face can be traced back to its original source.

🧭 Sources → Face Extraction → People → Reporting

DawaImg can ingest multiple data sources, extract faces, catalog them with provenance, merge them into People, and generate reports.
Sources Ingest post-capture data and streams 📁 Folders Nested dirs 📄 PDFs Pages/images 🌐 Web Downloads 🖼️ Image files JPG/PNG/etc 🎞️ Videos Frame extract 📷 CCTV stream Captured review Face Extraction Detect → crop → normalize Produces FaceRecords Catalog & Provenance Store links back to source files, pages, and frames SourceRecord → FrameRecord → FaceRecord Merge into People Cluster similar faces across all sources PersonRecord (investigator label) Person Reporting Exports, counts, timelines, and case summaries Reports from People All extracted faces remain traceable to their original source artifacts.
Notes:
  • “CCTV stream” here means captured footage reviewed post-capture (no live surveillance/alerts implied).
  • “People” are investigator-defined groupings of similar faces for organization and reporting, not identity claims.

Screenshots

DawaImg supports investigation-grade workflows. Here are a few screenshots.
Screenshot 1 Screenshot 2 Screenshot 3 Screenshot 4 Screenshot 5 Screenshot 6
🔌❌ Offline by design
No cloud services, no telemetry, no network dependency.
🧱 Built for case scale
Import entire folders and work with large image and video datasets. Designed to scale to millions of files.
🔍 Forensic review workflows
Post-capture analysis for investigations, documentation, and privacy-preserving collaboration.

🔗 Evidence Integrity & Chain of Custody

DawaImg uses a non-destructive processing model designed to preserve original data and document analysis steps. This helps private investigators and forensic teams keep review artifacts traceable back to the original case media.
Structured Record Model
  • SourceRecord – Created at ingestion to document the origin of imported data.
  • ResourceRecord – Generated per file to record file-level metadata and processing events.
  • FrameRecord – Created for each image, video frame, or PDF page extracted during review.
  • FaceRecord – Generated when a face is detected, linking directly back to its parent FrameRecord.
Each record maintains parent-child relationships, forming a traceable chain from derived artifacts (faces) back to original source material. Original files are never modified.
Non-Destructive Workflow
• Original media remains untouched
• All processing occurs on copies or derived artifacts
• Parent-child record linkage is preserved
• Results are reproducible within the same project environment
DawaImg provides structured provenance tracking for analytical workflows. It does not certify legal admissibility or replace formal forensic validation procedures.

🗂️ Files that can be searched

DawaImg can scan local and connected storage for supported file types and build a searchable index. Typical PI/forensics inputs include device exports, screenshots, social media captures, and investigative video.
🖼️ Images
Supported image formats:
.webp .bmp .png .jpg .jpeg .pbm .pgm .ppm .tif .tiff
Common inputs: screenshots, camera photos, device exports, OSINT captures.
🎞️ Videos
Supported video formats:
.mov .mp4 .avi .wmv .ts
Videos are processed via frame extraction during post-capture review.
📄 Documents
Supported document formats:
.pdf
PDFs are supported for embedded-image review and extraction.
Storage locations
Import and scan from:
  • Folders and nested directories
  • USB keys and external drives
  • Internal hard drives / SSDs
Also works with:
  • Mounted network shares
  • Remote connections (mapped/mounted storage)

👥 Is DawaImg a good fit for you?

✅ Good fit if you:
  • Are a private investigator, analyst, or part of a face/media forensics team
  • Work with large collections of case media (images, video, screenshots, PDFs)
  • Need automated face extraction and organization inside a single project
  • Require offline, local-only analysis (air-gapped or privacy-sensitive workflows)
  • Need repeatable review steps and exports with provenance to original frames/files
❌ Probably not a fit if you:
  • Need real-time monitoring, alerts, or live feeds
  • Expect cloud-based processing or SaaS workflows
  • Want preloaded face databases or identity-by-name lookup
  • Need the software to “verify identity” as an authoritative claim
  • Are looking for a consumer photo editor
DawaImg is designed for deliberate, post-capture analysis where control, documentation, and scale matter.

✨ DawaImg Features

🗂️ Extraction & Organization
  • Extract faces from still images
  • Extract faces from video (frame analysis) and group faces into “people”
  • Automatically cluster faces into distinct “person” groups
  • Organize people consistently across images, videos, and PDFs
  • Import and process entire folders and large datasets efficiently
  • Trace extracted faces back to original frames/pages and source files
🔍📊 Search, Review & Exports
  • Face and person search with configurable similarity thresholds
  • Case-media review workflows for large mixed datasets
  • Generate review outputs (exports) for follow-up and documentation
  • Group people by visual attributes (optional) to accelerate review
  • Video review: count and cluster faces across time windows
  • Project-based organization for keeping cases separated
🕶️ Privacy Tools
  • Face blurring for privacy and anonymization use cases
🧪🛡️ Labs & Defensive Research
  • Face recognition labs for evaluating robustness and failure modes
  • Defensive research into techniques that reduce face detectability
  • Design face masks to reduce face detection
  • Design face masks to reduce face recognition

⭐ Testimonials

Trusted in real-world workflows
DawaImg is used across investigation, journalism, and research teams that need offline, project-based face review.
Used by 400+ teams
Adopted by hundreds of investigation and analysis teams for offline face organization.
Used by PIs across the U.S.
Private investigators use DawaImg to triage mixed case media and export review outputs.
Used by U.S. news reporters
Reporters use offline workflows to compare faces across document releases and media sets.
Used by bloggers & research sleuths worldwide
Independent researchers use DawaImg for structured, reproducible analysis on local datasets.

💬 DawaImg – Frequently Asked Questions

1. What is DawaImg?
DawaImg is a free, offline face forensics organizer for macOS and Windows. It helps private investigators and forensic teams extract, organize, and analyze faces from large image/video datasets.
2. Is DawaImg really free?
Yes. DawaImg is free to use and does not require subscriptions, accounts, or online activation.
3. Does DawaImg include any face databases or sample data?
No. DawaImg ships completely empty. Users must supply all images and videos themselves.
4. Does DawaImg require an internet connection?
No. DawaImg operates 100% offline and has no cloud services, network dependencies, or telemetry.
5. Does DawaImg collect or transmit any data?
No. DawaImg collects no user data, no analytics, and no behavioral or usage metrics.
6. Can DawaImg be used anonymously?
Yes. There are no accounts, logins, licenses, or identifiers required to use the software.
7. Is DawaImg a surveillance or monitoring tool?
No. DawaImg does not support live feeds, continuous monitoring, alerts, or real-time tracking.
8. Does DawaImg support real-time face recognition?
No. All processing is post-capture only. Real-time or live face recognition is not supported.
9. Is DawaImg a government or law-enforcement tool?
DawaImg is an independent desktop application designed for private investigators, researchers, journalists, and analysts working with their own datasets.
10. Can DawaImg be used for doxing or harassment?
No. DawaImg is not designed to enable doxing, targeting, or harassment workflows.
11. How does DawaImg define a “person”?
A “person” is a software-generated grouping of similar faces. It does not represent a verified real-world identity.
12. Does DawaImg identify people by name?
No. DawaImg does not know who anyone is and makes no identity claims. It helps organize and compare faces within your dataset.
13. How accurate is face recognition in DawaImg?
Results are probabilistic and configurable via similarity thresholds. Outputs are analytical and should be validated within your investigation workflow.
14. Who owns the data processed by DawaImg?
You do. All data remains local, and users retain full ownership and responsibility for imported content.
15. Is DawaImg Open Source?
DawaImg is intentionally not open source. Releasing the code without constraints would make it easier to modify the software into harmful workflows beyond its intended scope.

📚 Research Datasets Processed with DawaImg

Download project bundles and open them in DawaImg to explore reproducible, post-capture workflows. All datasets listed below originate from publicly released sources. Extracted faces are unlabeled and carry no identity claims.
Dataset Download Description Scale Date Version
Public document release (House Oversight) Request access (verification required) Extracted faces from a publicly released document archive. Faces are unlabeled and intended for structured archive and review workflows.
Read Report 		~1,000,000 files · ~6 hours (local processing) February 4, 2026 1
Public document release (DOJ disclosures) Request access (verification required) Extracted faces from publicly released DOJ disclosure documents. Faces are unlabeled and structured for post-capture analysis. ~100,000 documents (public release) February 4, 2026 1
Minnesota public media archive project In progress · Contribute on GitHub Archive-grade organization of publicly available image and video material. Media is linked back to original sources and structured for review. No identity claims are made. Large-scale public media collection Dec, 2025 1
How to import projects
  1. Download the zip file.
  2. Unzip the file into a folder.
  3. Place the folder inside /home/DawaImg/projects/<Project Name>.
  4. Launch DawaImg; the new project will appear automatically.
  5. You can modify or export the project as needed.

🏆 Notable Uses

DawaImg is used in real investigations
  • Missing persons investigations (Florida and New York) — Used by licensed private investigators to organize and compare faces within user-supplied datasets during post-capture investigations.
    All analysis was conducted locally, with findings provided to appropriate authorities for follow-up.
  • Online image trace in a missing person case — Used to identify publicly available images associated with an active investigation. Analysis supported licensed investigators in connecting visual evidence to contact information for law enforcement follow-up.
  • High-profile missing person research (media use case) — Used by a newspaper reporter to compare source images against a large corpus of publicly available material during post-capture review.
  • Large-scale public event media archiving — Used to organize and catalog post-capture images and video from a publicly documented event for structured review and documentation.
  • Public archive documentation project (Minnesota) — Used to organize and structure large volumes of publicly available image and video content for journalistic review and documentation.

📨 Contact Us

For support, issues, and feature requests, use email or GitHub.
Email Support support@dawaforensics.com GitHub Issues & Updates