Redactor’s Latest Release 4.5 includes Image Redaction and Much More
March 27, 2023
share this article:
The Sighthound Redactor team is thrilled to share the latest major release of Redactor. This release delivers on one of the most requested features by now allowing users the ability to redact image files. There are additional features and fixes that increase usability and reduce the time necessary to complete redaction tasks. Redactor software administrators see some improvements regarding user password management. Read on for more about these enhancements.
Redactor now allows image files to be uploaded for redaction in addition to video files. Image files currently accepted include jpeg, jpg, png, apng, bmp, and jp2.
Previous enhancements regarding bulk selection and deletion features can be applied to video and image files at the time time. For example, you can have faces or license plates auto-redacted across several video and image files at one time.
The addition of image file support also brings updates to various buttons and menu options within Redactor. For example, when uploading video or image files, users will now click the “Add Media” button, previously seen as “Add Video”.
Redactor Release 4.5 User Highlights
Image File Export From Video
When reviewing pre-redacted or post-redacted video files, a user can now export a video frame as a png file. Just navigate to the desired time in the video and click the picture icon button by the timeline as shown below.
This can be very useful when needing to quickly share an image from a video and still be able to apply necessary privacy controls by blurring faces or obscuring location identifiers of the video.
File Names Retained After Upload via API
Media files imported via API will now display the filename on the Projects page as opposed to previously assigned ID names that included randomized alphanumeric characters. This change allows users to easily locate uploaded files using their internal naming conventions and maintain consistency during processing and exporting.
Adjustable Face Detection Box Size Configuration
The default face detection box size can be increased or decreased per installation of Redactor. The default setting is 1.5, which means that when the face is detected, the box around the face will be 150% the size of the face. If identifying characters such as hair, tattoos or other identifiable markers are outside of the box, they will not be redacted and may make identifying individuals from small groups easier. Increasing the box size may support improved privacy, but a manual review of redacted objects should always be performed to ensure the highest level of privacy is upheld.
Password Management Improvements
New password confirmation fields were added to the User management section and on the user’s “Reset Password” page. This should help to avoid incorrectly typed passwords by ensuring that entered password and confirmed password fields match before submission.
Administrator Feature - Recovery Mode
As administrators change for an organization, transition of administrative access is not always transitioned appropriately. With the new Recovery Mode, there is now an option to perform Redactor administration functions with a Recovery User. If you are in need of Recovery Mode directions, reach out to Redactor support.
System Logs Update for Enhanced Support
Redactor now features a dedicated log directory to store all log files, which can be zipped and sent along with support tickets. A new sysinfo.json file provides details on the user's computer specs (CPU, RAM, etc.) to further assist the support team.
Export Detection Metadata in JSON Format (Redactor API Only)
Redactor API users can now export detection metadata in JSON format, including each of the detected objects and where they appear in the image/video. This data can be used to provide a record of what was redacted or to validate redaction accuracy against a known dataset while testing different Redactor API parameters.
This release delivers on expanding Sighthound Redactor’s redaction application to include the redaction of faces, license plates, vehicles, and other special objects in images and not just videos. We are excited for our legacy and new customers to gain new efficiencies by working in Redactor to complete their redaction tasks quickly, efficiently, and accurately.
API users that process large quantities of files and have stringent privacy and reporting requirements are sure to appreciate the ease of exporting the detection metadata that is used to redact their video and image files.
Site administrators have more options to configure and manage their organization’s use and users. With more configuration options on the way, administrators should continue to share ideas on how to reduce Redactor management time and keep their team productivity high.