Pillar 03 · Creative Transparency System

Every image, honestly classified.

The Creative Transparency System (CTS) makes how an image was created visible to everyone — creator, voter, and buyer alike. Classification combines creator declaration with automated metadata checks, and conflicts route to a public review process.

Four categories

Every submission declares how it was made.

Camera Original (CO)

Unedited work straight from the camera. Verified by EXIF metadata and the absence of editing-software signatures.

Camera Enhanced (CE)

Photographs refined with traditional editing tools. Editing fingerprints from Lightroom, Photoshop or Capture One are expected and declared.

AI-Assisted (AA)

Human-authored work with meaningful AI contributions — generative fill, upscaling, style transfer. The human remains the author.

AI-Generated (AG)

Work where a generative model produced the primary image. Declared up front so voters and buyers always know what they're looking at.

How verification works

Declaration meets evidence.

Every upload runs through an automated detection cascade: a vision model checks for AI-generated and AI-assisted patterns, EXIF metadata identifies editing-software signatures, and the result is compared against the creator's declaration. Agreements are marked as verified. Conflicts route to moderation with a fast appeal for the creator.

📷 RAW Verified

The strongest authenticity signal we offer. Upload the original RAW (CR2/CR3, NEF, ARW, RAF, DNG, ORF, RW2, PEF, SRW) or HEIC/HEIF straight from the camera or phone, and we match its EXIF against the JPEG you posted. Same camera body, same capture moment, same body serial — verified.

EXIF + software signatures

Camera make, model, lens, capture timestamp, and editing-software signatures from Lightroom, Photoshop, Capture One, and generative editors. The primary signal for Camera Original vs. Camera Enhanced.

AI detection cascade

Every image runs through a vision model trained to spot AI-generated and AI-assisted patterns. The detected category is compared to the declaration; clean agreements pass, conflicts route to moderation, and the divergence is visible on the image’s detail page.

Community review

Reports from viewers feed a moderation queue. CSAM reports take a dedicated fast-path and are never routed to community reviewers.

Coaching Engine

Coaching for creators, not judges for contests.

Two tools extend CTS without compromising its integrity. The Coaching Engine is live today — immediate feedback on every upload so you can enter contests with confidence. The Edit-Loop comes in a later phase, keeping iterative work honest by construction so every iteration re-classifies itself.

Live · Coaching Engine

Immediate feedback on every upload.

The moment your image uploads, the Coaching Engine returns private feedback — composition notes, a technical score, suggested edits, and the contest categories that fit. No waiting, no queue. LensWideOpen is built to ensure your success: the feedback you need to compete with confidence, every time. Take the suggestions, stick with your original, or ignore the coaching entirely and go your own way — you stay in control of every decision. Critiques stay private — they never reach voters, leaderboards, or reputation — and run on a dedicated model with its own API key so they can never influence judging.

Later phase · Edit-Loop

Iterative edits that stay honest.

Apply a Coaching Engine suggestion — or your own edit — and the platform tracks the change. Each iteration auto-computes the new CTS category: a tonal adjustment stays CE, a sky replacement moves to AA, and crossing 30% cumulative regenerated area becomes AG permanently. Iterate as many times as you need to get the image contest-ready — the moment you enter a contest, that version is locked. No further edits, no way to edit and hide.

Classification is a feature, not a cage.

Every creative method is welcome on LensWideOpen — as long as it's honestly disclosed.