> ## Documentation Index
> Fetch the complete documentation index at: https://help.dingtalk.io/llms.txt
> Use this file to discover all available pages before exploring further.

# AI error detection for photo check-in

> DingTalk attendance photo clock-in AI recognition uses large Models and image understanding to automatically detect invalid photos such as black screens, blank walls, and missing faces, then generates error reports and pushes them to Admins for efficient, precise remote Supervised monitoring.

## Background and pain points

In daily attendance management, more and more organizations use photo check-in to verify whether employees have actually arrived on site or completed store inspection tasks. However, the following issues often occur in practice:

* Employees upload invalid images such as ceilings, floors, black screens, or blank walls.
* Photos do not include a face, so they cannot prove that the person was present.
* During store inspections, photos of shelves or displays are blurry or shot from the wrong angle and fail to reflect the real situation.
* Special job requirements (such as wearing a safety helmet or reflective vest, or complying with grooming standards) lack effective inspection methods.
* Admins must manually review hundreds or thousands of photos one by one, which is time-consuming, labor-intensive, and inefficient.

## Product solution

To address these management challenges, we launched the new AI error detection for photo check-in feature. Powered by **large Models and image understanding technology**, it automatically identifies non-compliant check-in behavior and helps Organization managers achieve efficient, accurate remote supervision.

### Key feature highlights

<CardGroup cols={2}>
  <Card title="Error image detection" icon="image">
    Automatically detect photos of abnormal scenes such as black screens, blank screens, ceilings, and floors.
  </Card>

  <Card title="Face detection" icon="face-viewfinder">
    Determine whether a photo contains a face.
  </Card>

  <Card title="Error notification mechanism" icon="bell">
    Automatically generate error reports and push them to Admins for handling.
  </Card>

  <Card title="Visual statistics Kanban" icon="chart-column">
    Provide an overview of photo quality data in the attendance console.
  </Card>
</CardGroup>

## Feature details and access path

<div style={{display: 'flex', gap: '12px', justifyContent: 'center', flexWrap: 'wrap', margin: '16px 0'}}>
  <img src="https://alidocs.oss-cn-zhangjiakou.aliyuncs.com/res/YdgOk2bv3LgBgq4B/img/d49dd7ba-40bb-4c67-968d-6649677ac213.png" alt="Feature details of photo check-in error detection" style={{width: '48%', minWidth: '240px', borderRadius: '8px', boxShadow: '0 2px 12px rgba(0,0,0,0.08)', margin: 0}} />

  <img src="https://alidocs.oss-cn-zhangjiakou.aliyuncs.com/res/YdgOk2bv3LgBgq4B/img/3776f58a-30d7-4c9b-8d22-89cb005849ab.png" alt="Photo quality data overview Kanban" style={{width: '48%', minWidth: '240px', borderRadius: '8px', boxShadow: '0 2px 12px rgba(0,0,0,0.08)', margin: 0}} />
</div>

* **Access path**: Attendance app → Statistics → More statistics → Check-in photo statistics.
* **Content**: On the page, review the analysis results for today's and historical check-in photos, including the total number of check-in photos, the number of qualified vs. unqualified photos, and the error type breakdown. Click View details to inspect each photo.
* **Generation time**: This feature currently runs in asynchronous post-event detection mode. Analysis typically completes and results are generated on the morning after check-in.

## Core value: higher efficiency, lower cost, smarter supervision

<CardGroup cols={3}>
  <Card title="HR / Admin" icon="user-tie">
    Cut manual review workload by more than 90% and quickly flag violations.
  </Card>

  <Card title="Business supervisor" icon="clipboard-check">
    Standardize process control for store inspections, field work, and on-site operations.
  </Card>

  <Card title="Organization-wide" icon="building">
    Strengthen policy enforcement as well as organizational discipline and safety management standards.
  </Card>
</CardGroup>

## Typical use cases

| Scenario                                  | Description                                                                                         | Representative industries             |
| ----------------------------------------- | --------------------------------------------------------------------------------------------------- | ------------------------------------- |
| Safety helmet check on construction sites | Automatically detect whether a safety helmet is worn during on-site check-in                        | Construction, engineering supervision |
| Kitchen hygiene cap compliance check      | Detect whether kitchen staff wear a hygiene cap when checking in for work                           | Catering, food processing             |
| Store employee grooming check             | Employees take a photo before starting a sales shift, and the system assesses grooming completeness | Beauty, chain stores                  |
| Product display compliance check          | Take photos of shelves during store inspections to determine whether displays meet standards        | Retail, fast-moving consumer goods    |