Automated Video Redaction Solution - Tender

Tender description

The Police and Crime Commissioner for Sussex on behalf of themselves and the Police and Crime Commissioner for Surrey ran a competitive flexible procedure to award a contract for video redaction. This had to be a cloud solution Amongst other things, the system had to be able to meet the following requirements: • Masking/blocking of logos, faces, vehicle registrations and any other features as identified by the user. • Ability to redact video with and without audio. • Ability to apply reverse redaction • The ability to apply zone redaction • Finalise the redaction • Progressive Editing enabled – ability to edit/redact over time rather than edit all in one go (e.g., edit over a couple of shifts) • Ability to review redactions and amend/tweak as necessary during the mark up stage. • Ability to use both automated (automatic tracking of faces – e.g., if person moves, need the mask to move with it and track it) and manual redaction options. • Comprehensive Redaction Editing Options (Frame-by-Frame, Real-Time Tracking, Audio, facial recognition, VRN (Vehicle Registration Number) recognition, skin blur, and object tracking) • Ability to upload various media types and convert redacted video files in industry standards formats such as MP4, AVI etc from the Surrey and Sussex DEMs platform (currently NICE Investigate) • Automatic recognition of other vehicle reg numbers • Ability to collate digital groupings of evidence (photos, videos, CCTV, etc.) into case folders • Redact audio that comes with images – e.g., remove speech other than the person requesting the info, or redact names and addresses stated by the person • Redact audio only footage - e.g., remove speech other than the person requesting the info, or redact names and addresses stated by the person • Run transcriptions from audio. • Transcription Feature for Audio Redaction with Keyword/Phrase Identification • Assign and Modify User Permissions for Each Case • Role-Based Access Control (RBAC) Implementation The first stage was a paper based exercise whereby suppliers self evaluated as to whether their solution met requirement as well as answering quality based questions. Those successful were taken through to stake 2 which involved live system demonstrations which were evaluated, The pricing was also assessed at this stage. The highest scoring supplier was taken through to stage 3 for testing, however upon evaluation the solution was not able to demonstrate an ability to meet those requirements which were contained within the tender. Therefore the second highest scoring supplier was invited to stage 3 whereby testing was successful and is now resulting in the contract being awarded. ## Contract award Title: Automated Video Redaction