Pimloc gets $1.8M for its AI-based visual search and redaction tool – ClearTips

Pimloc gets $1.8M for its AI-based visual search and redaction tool – TechCrunch

UK-based Pimlock has closed a £ 1.4 million (~ $ 1.8M) seed funding round led by Amedus Capital Partners. Existing investors Speedinvest and other anonymous shareholders also participated in the round.

2016-installed computer vision startup launches an AI Powerful photo classifier service named Folio in 2017 – providing service for smartphone users in such a way that they can recover the agency on their digital memories and transfer their data to cloud giants like Google.

The folio has since been ranked as an “expert search and discovery platform” for large image and video collections and live streams (such as those owned by art galleries or broadcasters) – and also a second tool powered by its intensive learning platform Launched. The product, Secure Redact, offers privacy-focused content moderation tools – enabling its users to search and reactivate personal data in visual content.

An example use-case that it gives is for law enforcement to anonymize body-cam footage, so it can be remodeled for training videos or prepared to be presented as evidence Can.

Pimloc CEO Simon Randall tells ClearTips that over the years, has been working with diverse image and video content to support businesses with classification, moderation, and data security challenges (image library, art gallery, broadcaster and CCTV provider).

“Through our work on the visual privacy side, we identified a significant gap in the market for services that allow businesses and governments to manage large-scale visual data protection on security footage. Pimloc As a result have worked in this field for some years to build capacity and product Pimloc Now business has been promoted around this mission. “

Secure Redact has two components: a first (automatic) phase that detects personal data (such as face, head, body) within video content. On top of this, Randall adds a layer of “intelligent tools” – allowing users to quickly review and edit results.

“All details and tracks are audible and editable before being accepted and reactivated by users,” he explains, adding that “personal data ID cards, tattoos, other objects and visual content, including phone screens (body cameras), and just faces in visual content Is wider than the wearer’s habit of taking messages on the phone’s screen, because they are typing, or a sensitive note on their laptop or notebook). “

A typical user mentioning the tool is the University of Bristol. An Associate Professor in Computer Vision, Dr. A research group, led by Dima Damon, is participating in an international consortium of 12 universities aimed at aggregating the largest datasets on Exogastric Vision – and being able to unfit video data sets for academic / open source use. Before providing for.

On the legal side, says Randall Pimloc Provides a range of data processing models – allowing for differences in how / where data can be processed. “Some customers are happy to act as data processors for Pimloc and use the Secure Redact SaaS solution – they manage their accounts, they upload footage, and before reactivation and use / editing Can review / update. Some customers run Secure Redact systems on their servers, where they are both data controllers and processors.

“We have over 100 users for the SaaS service who cover mobility, entertainment, insurance, health and safety. We are also in the process of setting up a host of on-premises implementations, ”he says.

Asked which areas Pimlock is developing the most for its platform in the coming years, it gives the following list: Security / analytics coming from the likes of Smart Cities / Mobility Platforms (Councils, Retailers, AVS With demand); The insurance industry, which he notes, is “using increasing amounts of visual data to monitor claims and risk” and thus “looking at systems responsible for data management and processing”; Video / telehealth, with the demand for driving into traditional consulting videos and driving for visual diagnosis; And law enforcement, where security goals need to be supported by “privacy designed by default” (at least where forces are subject to European data protection).

On the competitive front, he notes that startups are increasingly focusing on specialist application areas for AI – arguing that they have the opportunity to create compelling end-to-end proposals for large tech companies Are difficult.

In particular for Pimlock, he argues that his particular security-focused niche has an edge – given “Deep Expertise” and Distinct Domain Experience.

He added, “There are low barriers to entry to create low-quality products, but high technical barriers that are enough to create a service that is large enough to be used extensively with wild ‘footage in real’ is.” The generalist services of the big tech players do not coincide with the domain specific provisions of Pimlock / Secure Redact. Video security footage for AI is a difficult domain, with systems trained on lifestyle / celebrity or other common data sets performing poorly on actual security footage. “

Alex van Somren, MD of Amadeus Capital Partners, commenting on seed funding in a statement Said: “Privacy is a key requirement by design and large-scale solutions, as video grows as a data source for mobility, insurance, commerce and smart cities, while our dependence on video for remote work increases. is. We are very excited about the ability to PimlocProducts to meet this challenge.

“Consumers around the world are concerned about how enterprises are handling the increasing amount of visual data being captured 24/7. we believe Pimloc Has developed a leading industry for visual security and privacy that will allow businesses and governments to manage the use of visual data while protecting consumers. We are excited to support their vision as they expand into broader enterprise and SaaS markets, ”said Ricking Hao, principal at SpeedInvest, in another supporting statement.

Be the first to comment

Leave a Reply

Your email address will not be published.