Imaging pioneer delivers AI-driven face recognition and visual intelligence to volume-portrait market
photobook.ai, developer of AI-powered photobook technology, will debut industry-leading AI solutions at the SPAC 2019 annual school and sports photography conference in Las Vegas, NV. The company will be exhibiting applications of its Albumstory AI-SDK Jan. 18-19 in booth #310.
Albumstory AI-SDK offers developers and brands industry-leading imaging AI, computer vision and face detection plus recognition. Most importantly for sports, school and event photography companies, Albumstory is able to perform all of its imaging magic on the device, not the cloud, making it a truly convenient solution for location-based shoots.
“Our conversations with leading volume photography firms have already shown that Albumstory AI-SDK can improve photo selection and organization on-site,” says Terence Swee, CEO and founder, photobook.ai. “We’re excited to show how Albumstory can improve workflow, make possible new product categories and boost profits for SPAC attendees.”
The Albumstory AI-SDK is organized into four major components:
- iFace SDK is an advanced face intelligence engine with face detection, face recognition and face-quality algorithms. The iFace SDK, when used together with Curation and the Events Engine supports storytelling based on people, like creating a Collection of individual players on a team from across multiple sets of photos over multiple years. The proprietary face-recognition technology works in shadows, with eyewear, slight head yaws, head and neck wear and with other partial occlusions.
- The Events Engine uses a proprietary Events Intelligence algorithm which creates Collections based on time and place, things and/or frequently occurring faces, hence automatically proposing meaningful stories worth curating.
- The Curation Engine contains the AI, semantic and computer vision algorithms to analyze the Collection. It ranks images by its quality, removes duplicates, ensures visual variety in the mix, detects faces, removes documents and identifies key images like those of food, coffee, landscapes, people, cats, dogs and babies. Its breadth of vocabulary and understanding of the world is constantly expanding as it uses deep learning neural networks to continuously refine and learn to recognize often-captured moments by users.
- The Layout Engine uses intelligent design rules to create layouts to target page and photo dimensions. This engine can choose a single photo from a burst series, automatically place panoramas as a spread and avoids cropping faces regardless of layout. The Layout Engine can also be totally replaced by the auto-fill and templates of licensees whereby it can be asked to only return a suggested page-grouping list.
From its deep bench of talent and software stacks, photobook.ai is looking for strategic partners from the school and volume photography space to closely use its deep learning engines to recognize and curate photo sets for volume photography scenarios. For example, it is already working on automatically detecting and understanding sports jersey numbers of players, which are really useful for team sports which requires a helmet, making face recognition a non-viable option.
“Now licensees can collect an album of all photos of Ethan, in jersey number 23 across all three past seasons,” says Swee. “Automatically. Photobook.ai wants to help the industry use AI to radically monetize their large volumes of photos.”
Press and media members can arrange to meet Terence at SPAC 2019 using this link: https://terence-photobook-ai.youcanbook.me.