Practical guides for finding similar images, pose-based search, filtering by color, removing duplicates, and finding people.
From version 3.14, Assist connects to WorkBuddy. Describe a picture or attach a reference image in the conversation to search your local library, narrow the scope, filter results, and see the pictures in the chat.
Read cycle usage, reset dates and daily details, and distinguish image traffic from other requests and client seats.
Preview keypoints and heatmaps, tune feature limits and distribution, and understand saved settings and existing-library reprocessing.
Compare two images directly using global, subject, color and local-keypoint strategies without importing them first.
Prepare a library, choose and edit a reference, select the right strategy, review results, and organize or export single and batch searches.
Parse faces, limit grouping to an event folder, review the clusters and export folders for each group.
Import embedded Excel images, search for matching lights, inspect worksheet and quotation details, and export a verified shortlist.
Prepare local-keypoint data, crop a distinctive detail, search with Local · Keypoints and verify the complete image in the results.
Find candidate portraits without a reference face using parsed face attributes, search scope and result limits.
Find similar body positions using a photo or an editable skeleton, then control orientation, framing and search scope.
Connect a phone or tablet to the host, sign in as a client member, search and browse images, and troubleshoot connection problems.
For brand colors, same-tone material, and visual-style organization, view dominant colors, use color search, combine color labels and list filters, and export results for designers.
For new computers, new drives, or old material libraries, back up the database, migrate the database location, handle missing paths, confirm statistics, refresh source paths, and repair database.db only when damaged.
When image, pose, semantic, face, or color search has no results, troubleshoot database statistics, task progress, search scope, filters, strategy data, and default settings step by step.
Create a project-specific collection, add results from multiple searches, filter and browse within the collection, manage naming, and export a deliverable material package.
For events, reception, or portrait checking, choose a USB camera, configure search scope and frequency, review live results, capture photos, and verify hit records.
For many reference images at once, enter batch search, choose search type and scope, review each row of matches, expand more results, and export the whole matching set.
For event, client, employee, or portrait libraries, generate face data, use face search and face albums, set scopes, review results, and export photos by person.
For screenshots, poster crops, or revised design images, use paste search, cropping, background removal, color strategy, and result filtering to recover the original image or find similar material.
For folders containing images, Excel, PDFs, videos, screenshots, and subfolders, prepare the environment, add the root folder, wait for parsing, review extracted images, add search data, and refresh later changes.
Turn an unclear image request into subject, scene, color, and use-case terms, run semantic searches, narrow with scope and filters, collect candidates, and export deliverables.
Enable duplicate detection, prepare duplicate data, choose a database, folder, or collection scope, review duplicate groups, and distinguish between removing, deleting, and removing from collections.
A maintenance workflow covering database status checks, refresh after file changes, index rebuilds when search is abnormal, task progress, backup or migration, and repair for damaged database.db files.
Use one reference image to search similar material, crop the target when needed, limit the search scope, review results, and add matches to a collection or export them for delivery.
A complete first-use workflow: enter the workspace, handle database and spreadsheet prompts, add a material folder, watch indexing progress, confirm database statistics, and run your first similar-image search.
Prepare image features, choose a clustering mode, adjust groups and export the current grouped images.
Distinguish the system, database and export drives, free the right space, and preserve originals and database files.
Compare face pairs from two photos and understand face selection, scores and thresholds.
Turn unnamed face groups into a reusable people list, review assigned faces, correct mistakes and distinguish photo counts from face counts.
