Reviewing AI Detections

After SuperClear AI runs through your photos, the SuperClear AI Screen shows each photo with the AI's detected defects marked as bounding boxes and labels. This article walks through everything on the screen — the toolbar, filters, photo cards, status indicators, image editor, and bulk actions — so you can confidently review and refine the AI's results.

Opening the Review Screen

1. Sign in to superclearai.nspec.app
2. From the dashboard, find your inspection
3. Click View SuperClear AI on the inspection card (or open the inspection and click "Open SuperClear AI Inspection" from the Detailed Inspection View)
4. The SuperClear AI Screen opens with the photo grid

The Header

At the top of the screen you'll see:

• Back arrow — Returns to the Detailed Inspection View
• Project name and inspection details — Inspection name, visit number, date, inspector
• Stats line — "X Photos · Y Defects · Z In Progress" (the In Progress count turns red/orange when photos are still processing)
• Logout button — Right side, with confirmation dialog

The Toolbar

Below the header is the toolbar, which has different controls depending on whether you have any photos selected.

Left side

• Upload Photo (green, upload icon) — Opens the photo upload modal to add more photos
• Report (blue, document icon) — Generates a PDF punch list. See Creating Punch Lists from AI Detections.
• Generate AI Report (blue, lightning icon) — Runs the AI to write captions, observations, and action items into the inspection. See Generating AI Reports.
• Download QA Report (blue, download icon) — Appears after an AI report run finishes; downloads the QA field report as a PDF. See Generating AI Reports.
• Select All checkbox — Quickly select every visible photo
• Selection counter — Shows "X of Y selected" when photos are checked

Middle (when photos are selected)

• Regenerate AI (green) — Re-runs AI detection on the selected photos. See Regenerating AI Analysis.
• Cancel AI (orange, only visible if a batch is currently running) — Stops in-progress AI processing
• More menu (three dots) — Includes Download (zip the selected photos) and Delete Selected

Right side

• Settings menu (gear icon) — Opens Photo Settings (confidence threshold, condition colors) or Edit Project

The Filters Panel

Below the toolbar is the Filters panel, which is collapsible. Click Filters to expand it.

• Only Show Defects toggle — Hides photos that don't have any AI-detected defects
• Group by Conditions toggle — Organizes photos into collapsible groups by condition tag instead of a flat grid
• Filter by dropdown — Multi-select to focus on specific conditions (parent tags only)
• Sort dropdown — Upload Order, Annotations, or Tags
• Direction button — Toggle ascending (↑) or descending (↓)
• Reset button — Clear all filters back to defaults

The Photo Grid

The main area of the screen is the photo grid (or grouped view if you toggled Group by Conditions). Each photo is shown as a card with the AI's annotations baked into the image.

Anatomy of a photo card

• Image — The photo with bounding box annotations overlaid for any AI-detected defects
• Checkbox (top-left) — Select for bulk actions
• Location badge (top-right) — "Building · Floor · Elevation" if metadata is set
• Order badge (bottom-right) — Sequential photo number
• Action overlay (appears on hover) at the bottom of the image:
  • Edit (blue) — Opens the image editor
  • Delete (red) — Removes the photo with confirmation
  • AI Detect (green) — Re-runs AI on this single photo
• Status circles (below the image) — Show AI processing progress for each step
• Caption — Editable text under the photo
• Photo Conditions button — Opens a dropdown to assign tags

Status circles

Each photo card shows three status circles representing the AI processing steps:

1. Step 1: Conditions — Classify what kind of WRB defect (if any) is in the photo
2. Step 2: Detection — Detect the specific defects and produce bounding box coordinates
3. Step 3: Drawing — Render the bounding boxes onto the photo so you see the defect markups

Each circle shows one of these states:

• Pending (gray) — Waiting to be processed
• Processing (animated spinner) — Currently being analyzed
• Completed (green check) — Done; click to view the AI's reasoning
• Error (red X) — Processing failed; click to see why

Click any completed circle to open the Analysis Details modal, which shows the AI's notes about that step. The modal includes a disclaimer: "This analysis was generated by AI and may not be fully accurate. Always verify results with professional judgment."

Grouped View

Toggling Group by Conditions in the filters panel reorganizes the grid into collapsible sections by condition tag. Each section shows:

• Condition name and photo count
• Expand/collapse chevron
• Select All in Group checkbox
• Download Group button

Use grouped view when you want to focus on specific defect types — for example, reviewing all "Tape Failure" findings before moving on to "Fastener" findings.

Editing a Photo

Hover over a photo and click Edit in the action overlay to open the Image Editor. The editor is a full drawing tool with seven annotation types, a color picker, stroke slider, zoom, undo/redo, and a visibility filter. For the complete reference — every tool, every keyboard shortcut, and best practices — see Annotating Photos with the Image Editor.

Editing Captions and Metadata

1. Click anywhere on a photo card to enter edit mode
2. Edit the caption directly in the textarea
3. Set Building, Floor, and Elevation from the dropdowns (only available if the project has Building/Floor Options configured)
4. Click the green checkmark to save, or the red X to cancel

Bulk Actions

1. Check the boxes on the photos you want to act on (or use Select All)
2. Use the toolbar buttons to:
   • Regenerate AI — Re-run AI detection on the selected photos
   • Download (in More menu) — Zip and download the selected photos
   • Delete Selected (in More menu) — Permanently remove with confirmation
   • Generate Report — Punch list with only the selected photos

While AI Is Processing

When a photo is being analyzed, you'll see a scan-line animation overlay on the image with corner decorations and the text "Analyzing...". You can't interact with that photo until processing finishes. The status circles update in real time as each step completes.

Troubleshooting

The AI didn't find any defects

Solutions:

• The photo may not contain any detectable WRB defects — that's a valid result
• Try lowering the confidence threshold in Photo Settings to see lower-confidence detections
• If you're confident the AI missed something, use the image editor to add your own annotation
• Make sure the photo quality is good — see Photo Best Practices

The AI flagged too many things

Solutions:

• Raise the confidence threshold in Photo Settings to filter out lower-confidence detections
• Use the image editor to delete incorrect markups individually

Photos are stuck on "Analyzing..."

Solutions:

• Wait — AI processing for large batches can take several minutes
• Check the In Progress counter in the header — if it's not changing, the batch may have stalled
• Use Cancel AI in the toolbar to stop a stalled batch, then Regenerate AI to retry

Next Steps

• Managing Photo Tags and Conditions
• Understanding AI Detection Scores
• Regenerating AI Analysis
• Creating Punch Lists from AI Detections
• Managing Multiple Inspections

---

Need help? Contact our support team at [e-mail]