Skip to content

Developing Work Skills · Career & Development

Working with Images, Audio and AI

Build a reviewable image-description or transcript workflow with permitted media and editable text.

Level
Advanced
Estimated study
About 12 hours
Lessons
9 lessons in 3 modules
Format
Online, self-paced
Start your free 7-day trialAlready a student? Open course

Monthly: $0 today with a card, then $34 on day 8 unless you cancel before then. After the trial, 30 days' notice applies. Active on-campus students have academy access included. Check your existing student account before buying another subscription.

About this course

Extend an AI application with permitted image or audio inputs. Practice consent, accessible alternatives and checks that help a person review uncertain model output.

What you will be able to do

  • Validate permitted media inputs and explain their privacy implications.
  • Build an image-description or transcription workflow with server-side credentials.
  • Measure errors and provide an editable text result and accessible alternative.

Course outline

Module 1: Choose and validate media

3 lessons · 4 h

Check permission, file types, size limits and retention for sample inputs.

  1. Choose permitted media and a retention policy1 h 20 min
  2. Validate type, size and decode limits1 h 20 min
  3. Minimize metadata and preserve useful provenance1 h 20 min

Module 2: Build a media workflow

3 lessons · 4 h

Request a description or transcript and handle provider errors and usage costs.

  1. Request a bounded image description1 h 20 min
  2. Request a transcript without inventing speech1 h 20 min
  3. Control failures, duration and usage costs1 h 20 min

Module 3: Review uncertain output

3 lessons · 4 h

Check representative samples, allow corrections and provide accessible text.

  1. Measure errors against representative samples1 h 20 min
  2. Offer corrections and accessible text1 h 20 min
  3. Deliver the reviewable media assistant1 h 20 min
Get helpStart 7-day trial