Capabilities
Kid Smart AI leverages a range of open-source and proprietary technology to build cutting edge AI solutions for industries who build products for children, like education, toys and games, and content generation.
| Capability | Description | Endpoint |
|---|---|---|
| Verbal fluency assessment | Assesses the fluency of verbal literacy for any given text. | /v1/audio/fluency |
| Pronunciation assessment | Evaluates the accuracy of pronunciation for spoken language. | /v1/audio/pronunciation |
| Word Recognition | Evaluates accuracy of spoken words/phrases (up to 30 seconds). | /v1/audio/recognition |
| Decodable fictional stories | Generates personalized fictional stories for readers. | /v1/text/decodable-stories |
| Image generation | Creates personalized and consistent character illustrations. | /v1/image/character-generation |
Verbal fluency assessment
The Verbal fluency assessment product provides fast, accurate analysis of a fluency assessment of a child. The student is recorded reading a passage, then that recording and an ID for the text that was read are passed to Kid Smart AI API.
Inputs:
- Audio file (mp3, mp4, etc, specified here)
- ID of text that was provided to the child
- The texts must be provided by the organization, or they can use any Kid Smart AI text
Outputs for Verbal Fluency Assessment
| Output Description | Details |
|---|---|
| Words per minute (WPM) | Measures how many words the child reads per minute. |
| Accuracy of reading | Percentage of words read correctly (calculated as 100 * (Number of words correctly read / total number of words)). |
| Errors identified | Classifications include Pronunciation, Omission, Insertion, Self-correction, Repetition. |
| Timestamps of errors | (Optional) Provides the specific times at which errors occurred during the reading. |
The analysis occurs after the submission of the audio file, typically within 30 seconds of submission.
Best practices
- In noisy environments (like a typical classroom), please use a headset for the audio collection. a. If you cannot make out the words from an audio recording, neither can Kid Smart AI
- Allow the child the opportunities to self correct
Pronunciation Assessment
The Verbal fluency assessment product provides fast, accurate analysis of a fluency assessment of a child. The student is recorded reading a passage, then that recording and an ID for the text that was read are passed to Kid Smart AI API.
Inputs:
- Audio file (wav only, specified here)
- Expected Phoneme or phonemes (this can be in Arpabet, or in your own phonenic alphabet)
- Discuss with Kid Smart AI your phonemes, we can implement it.
- Model ID
- We can build custom model outputs for your use case, or you can select from one of our options based on confidence (high vs medium) and phoneme segmentation requirements (none vs segmented)
- Please see playground examples to see the differences.
Outputs for Phoneme assessment
| Output Description | Details |
|---|---|
| Is Correct? | Boolean indicating whether the phonemes were correctly pronounced |
| Analysis details | Timestamps and the predicted phonemes. If there is uncertainty in the phoneme prediction, multiple phonemes are given |
| Feedback | Feedback any errors that were identified what was identified |
The analysis occurs after the submission of the audio file, typically within 30 seconds of submission.
Decodable story generation (coming soon)
Inputs:
- Select a scope and sequence (select from ours, or provide your own)
- Specify the level on the provided scope and sequence
- Specify topic of story
- Specify length of story
- Optional: Specify any skill that should be used more often, default is the the level
- Optional: Specify additional words to use
Outputs: Decodable text using only the vocabulary provided of the specified length