AI Kit add-on allows for an effective management of assets and improves searchability of the library:
Image Intelligence component provides a full-fledged set of features for image processing.
Video Intelligence component includes various video analyzing tools— one of which is the Transcription feature.
Transcription is an innovative feature that allows you to convert speech from video or audio into text. With this new feature, Pics.io users can easily obtain accurate transcripts of their recorded content, making it much easier to reference, search and share videos and audio files.
NB! Audio transcription is currently available for the following formats: 'audio/mpeg', 'audio/mp4', 'audio/x-wav', 'application/ogg'.
Enable/disable access for your teammates
To restrict certain members of your team from generating captions for your video and audio assets or from access to the functionality as a whole, you can configure permissions for their roles in the Roles section.
For that, select a specific role, open Collection permissions, select the necessary collection/s, and navigate to Assets > Change asset attributes, and tick the checkbox Transcription.
Convert speech to text
The Transcript feature can be accessed in the Transcript panel or previewed in the Details panel. To transcribe your video or audio file, follow the next steps:
1. Specify the language
Choose the one you want transcripts to be generated in from the list of languages. It should be the one spoken in the video or audio asset.
NB! If you change the language for an existing transcript, a confirmation modal will ask if you want to regenerate the transcript.
2. Transcribe
Click the sparkles icon to start the process. First results usually take up to 1,5 minutes to be displayed and the transcription status (Processing) can be tracked within the Details panel at the top.
When the work is done, the time codes with the text script will appear in the Transcript panel next to the Analytics.
Tip: You can use a transcript time code to jump to a specific moment in the video or audio.
Speakers & Badges
The Transcript panel now includes speaker identification, making it easier to follow conversations in your video or audio files. Each detected speaker is displayed next to the corresponding timestamp with a speaker badge.
By default, speakers are labeled sequentially (e.g., Speaker 1, Speaker 2, Speaker 3), but you can easily rename them by hovering over a badge and clicking the pencil icon. Once renamed, the new speaker name is automatically applied across all related transcript lines.
To improve readability, the transcript text is also grouped by speakers, both in the default view and in reader mode.
Search words and phrases
Insert a word in a search field to filter out the results and display only those that contain the highlighted word. The results act as timestamps allowing you to jump to the selected point within a video or audio timeline and play it further.
NB! If you perform a search by transcript in the filter panel, the operators Starts with and Ends with work by transcript blocks that are divided by timestamps. The operator Contains doesn't work with several words that are located in different transcript blocks.
For example, if you search with the operator Contains by the phrase "To start off", it won't find this video since words are located in separate blocks.
Copy or edit transcript
To make changes to the existing line of converted speech, mouse over the necessary timecode and click the pencil icon to edit the text. Changes will apply automatically upon leaving the field.
You can also copy transcript text with a single click - just hit the “Copy” button to grab the entire transcription instantly.
Turn on subtitles
The subtitle feature is enabled by pressing the Subtitles (CC) button in the menu below the timeline of your video or audio file. The video/audio and transcript are synchronized meaning you can use either of them to jump to a specific moment in the asset.
Pro tip: Subtitles are available when an asset is shared externally or added to a shared collection (website), meaning you can share your transcribed videos with people who do not have direct access to your Pics.io library using a single link and they will be able to watch/review video assets with subtitles.
Learn more about single asset sharing and or sharing assets in customized collections (websites).
Download transcripts in SRT
The converted speech can be downloaded in SRT file format which is a text format used for subtitles. The file will contain time codes with transcripts.
Automate speech conversion
Check the option and choose a default language in the Artificial Intelligence section of My team settings to allow all newly uploaded video and audio assets to be recognized with transcription.
Pricing
Image-related features like AI keywording, Face recognition, and Image Captioning utilize the AI credits.
On the other hand, video processing features, such as Transcription, use the AI minutes.
AI Kit provides a different amount of AI credits or minutes depending on the AI Kit Tier selected. You can see the amounts on the Billing page and choose whichever AI Kit Tier suits your needs.
In case you cannot find a fitting AI Kit Tier that would contain the amounts you need, be sure to contact us at support@pics.io and we can can work it out.
Transcription is available as a part of AI Kit's Video Intelligence package. If you’re using the legacy AI Kit add-on, nothing changes — transcription remains available there as long as you keep your current subscription.
Feel free to refer to the AI Kit article for easier comparison of newer and legacy versions.
You might also be interested in: