Periodic Reporting for period 2 - Didimo (Humanity in Digital Communication)
Período documentado: 2020-10-01 hasta 2021-09-30
Nowadays, every minute, people across the world are making billions of virtual connections – for work, for entertainment, for information and for social communication. However, most of these interactions are still missing core human attributes. A 3D virtual self improves everyday interactions online for all users (facilitates closer online connections, user experience improvements and avoids digital isolation). Our mission is to project humanity within that world and to bring the richness of human communication to every online interaction settled on the following value propositions:
-Authenticity: We allow people to express their internal authentic persona and to access worlds and experiences they wouldn’t otherwise be able to;
-Adaptability: Our models can be deployed in any 3D environment. We offer the possibility for Customers to use their own rig and mesh or to use ours;
-Scalability: We automate a complex series of steps usually performed by teams of professionals and expensive equipment, enabling a global and diverse reach;
-Simplicity: We deliver an effortless UX. Users can create high-fidelity avatars in a few simple steps across a growing range of partner Apps;
-High-Fidelity: Our patented technology is backed with cutting-edge research to create realistic and close to human interactive models.
Didimo has upscaled Didimo technology to Technology Readiness Level (TRL)8 towards a complete and qualified solution ready for commercialisation. Two pilots - in mobile and fashion retail - were conducted to test two different market segments. Didimo has launched the Didimo Showcase app for iOS and the high-fidelity improvements to the didimo assets for a more realistic and authentic display. Together, these pilots helped to keep addressing our overall mission - humanizing technology - of putting a didimo into the hands of every consumer (through their mobile phones or other digital devices) and capitalising on their value across multiple platforms and sectors.
-Authenticity: We allow people to express their internal authentic persona and to access worlds and experiences they wouldn’t otherwise be able to;
-Adaptability: Our models can be deployed in any 3D environment. We offer the possibility for Customers to use their own rig and mesh or to use ours;
-Scalability: We automate a complex series of steps usually performed by teams of professionals and expensive equipment, enabling a global and diverse reach;
-Simplicity: We deliver an effortless UX. Users can create high-fidelity avatars in a few simple steps across a growing range of partner Apps;
-High-Fidelity: Our patented technology is backed with cutting-edge research to create realistic and close to human interactive models.
Didimo has upscaled Didimo technology to Technology Readiness Level (TRL)8 towards a complete and qualified solution ready for commercialisation. Two pilots - in mobile and fashion retail - were conducted to test two different market segments. Didimo has launched the Didimo Showcase app for iOS and the high-fidelity improvements to the didimo assets for a more realistic and authentic display. Together, these pilots helped to keep addressing our overall mission - humanizing technology - of putting a didimo into the hands of every consumer (through their mobile phones or other digital devices) and capitalising on their value across multiple platforms and sectors.
In the DIDIMO Project, the research and development (R&D) activities focused specifically on two sectors – Mobile and Fashion Retail.
For the Didimo Mobile product, we were committed to involving multiple mobile global market leaders which are currently Didimo customers. In late 2020, Didimo launched the updated platform Customer Portal and Didimo Showcase – this last one was recently rebranded to Didimo Showcase. This launch included the release of Didimo’s Android app, Unity SDK and Unity Editor Plug-in. In this announcement which targeted nearly 24k recipients, Didimo presented it as “the first low-cost, high-fidelity, fast-turn-around solution on the market - automating digital human generation in around 90 seconds from a single selfie”. Our technology – new app and Unity tools (SDK and Plug-in) – provides the user with an increased level of quality and ease to integrate digital humans into third-party projects. With enhanced visual quality, improved speed, and added features like a new hair library, customers can fully configure their didimos for the most authentic and impressive experience in their applications. Didimo Showcase app (for iOS) provides the user with an opportunity to create avatars with an initial selfie from a smartphone. UX and UI improvements on the app, including a redesigned gallery, intuitive options to generate and edit a didimo, when matching with a greater visual quality-display of didimos will enable an easier and pleasant experience, becoming a force against churn rate of users and expanding our entry door to new customers.
The Didimo Fashion Retail product was developed for e-commerce organisations to find a solution to how clothes fit and look on their customers. Initially developed for Farfetch for an immersive virtual try-on (VTO) at scale on an online retail store for luxury fashion, the Didimo Fashion Retail product is for an online game application - DREST, where users have the opportunity to order the physical version of luxury clothes that they virtually try on within the mobile game. The integration of Didimo technology offers a realistic representation of a photorealistic model ready to be dressed, represented by a high-fidelity didimo. For the first version of the Pilot Fashion Retail product, different tests were performed around the player to illustrate and interact with the model - a full body avatar - by editing zoom views and changing from different camera positions. The VTO displays a didimo in a customisable full body in a 360 degree visualisation. For the second version of the Pilot Fashion Retail product, technical improvements on accessibility and usability of the first version were performed. This new version included an improvement in the efficiency and security methods on asset caching and authentication renewal respectively, as well as new camera controls and displays. A bespoke development on Didimo technology was made to better match Drest needs on face representation as well as to offer a realistic full-body representation. A group of meshes have been designed and refined for DREST, including the following new options: lipsticks and lip liners, eye lashes, eyeliner and shadows, blush and face mesh.
For the Didimo Mobile product, we were committed to involving multiple mobile global market leaders which are currently Didimo customers. In late 2020, Didimo launched the updated platform Customer Portal and Didimo Showcase – this last one was recently rebranded to Didimo Showcase. This launch included the release of Didimo’s Android app, Unity SDK and Unity Editor Plug-in. In this announcement which targeted nearly 24k recipients, Didimo presented it as “the first low-cost, high-fidelity, fast-turn-around solution on the market - automating digital human generation in around 90 seconds from a single selfie”. Our technology – new app and Unity tools (SDK and Plug-in) – provides the user with an increased level of quality and ease to integrate digital humans into third-party projects. With enhanced visual quality, improved speed, and added features like a new hair library, customers can fully configure their didimos for the most authentic and impressive experience in their applications. Didimo Showcase app (for iOS) provides the user with an opportunity to create avatars with an initial selfie from a smartphone. UX and UI improvements on the app, including a redesigned gallery, intuitive options to generate and edit a didimo, when matching with a greater visual quality-display of didimos will enable an easier and pleasant experience, becoming a force against churn rate of users and expanding our entry door to new customers.
The Didimo Fashion Retail product was developed for e-commerce organisations to find a solution to how clothes fit and look on their customers. Initially developed for Farfetch for an immersive virtual try-on (VTO) at scale on an online retail store for luxury fashion, the Didimo Fashion Retail product is for an online game application - DREST, where users have the opportunity to order the physical version of luxury clothes that they virtually try on within the mobile game. The integration of Didimo technology offers a realistic representation of a photorealistic model ready to be dressed, represented by a high-fidelity didimo. For the first version of the Pilot Fashion Retail product, different tests were performed around the player to illustrate and interact with the model - a full body avatar - by editing zoom views and changing from different camera positions. The VTO displays a didimo in a customisable full body in a 360 degree visualisation. For the second version of the Pilot Fashion Retail product, technical improvements on accessibility and usability of the first version were performed. This new version included an improvement in the efficiency and security methods on asset caching and authentication renewal respectively, as well as new camera controls and displays. A bespoke development on Didimo technology was made to better match Drest needs on face representation as well as to offer a realistic full-body representation. A group of meshes have been designed and refined for DREST, including the following new options: lipsticks and lip liners, eye lashes, eyeliner and shadows, blush and face mesh.
As a core component of our Platform, Didimo is continuously monitoring the API performance to return an efficient integration of didimos into the third-party Clients. Didimo, with Didimo API, aims to offer a powerful, easy, secure and swift generation of realistic didimos.
The pipeline is accessible through the API and encapsulates all stages of the production pipeline: model, rig, animation, and render.
The future developments identified were:
-Updates on the didimo representation as we continue to make likeness improvements.
-Developments on the Pipeline to include improvements on the overall generation process, a body mocap from iPhone, rendering in real-time, and others.
-Developments in the detection of age, skin tone and gender, hair removal and volume calculation to fit the hair is a further and isolated step and improving the skin segmentation method.
-Developments for Didimo Showcase on motion capture app for body integration and addition of body capture capabilities to the mocap app.
The DIDIMO project impacts during the second year were:
-Didimo effectively supports organisational development and scalability.
-Conduction of training and capacity building activities for the development needs of the team.
-Didimo has recruited a total of 22 candidates to join the team.
-Didimo has extended its Customer Portfolio with the commercial contract made with establishing a large enterprise deal with a new retail customer.
-Didimo market has internationally extended by winning the acceleration program Portugal to Take Off which has provided mentoring and leads for the US market.
-News about the company’s accomplishments, new team members, funding rounds, hiring goals and others were created and disseminated.
-The relationship between Didimo and the media was reinforced, with interviews and opinion article opportunities. A total of 67 press releases were distributed and articles were published in the media.
-Didimo created content about humanizing technology and ethics for digital humans and disclosed it to position Didimo’s spokesperson as an opinion maker on this topic.
-Didimo has increased its social presence in the second year, increasing the number of followers from 1800 to more than 2700.
The pipeline is accessible through the API and encapsulates all stages of the production pipeline: model, rig, animation, and render.
The future developments identified were:
-Updates on the didimo representation as we continue to make likeness improvements.
-Developments on the Pipeline to include improvements on the overall generation process, a body mocap from iPhone, rendering in real-time, and others.
-Developments in the detection of age, skin tone and gender, hair removal and volume calculation to fit the hair is a further and isolated step and improving the skin segmentation method.
-Developments for Didimo Showcase on motion capture app for body integration and addition of body capture capabilities to the mocap app.
The DIDIMO project impacts during the second year were:
-Didimo effectively supports organisational development and scalability.
-Conduction of training and capacity building activities for the development needs of the team.
-Didimo has recruited a total of 22 candidates to join the team.
-Didimo has extended its Customer Portfolio with the commercial contract made with establishing a large enterprise deal with a new retail customer.
-Didimo market has internationally extended by winning the acceleration program Portugal to Take Off which has provided mentoring and leads for the US market.
-News about the company’s accomplishments, new team members, funding rounds, hiring goals and others were created and disseminated.
-The relationship between Didimo and the media was reinforced, with interviews and opinion article opportunities. A total of 67 press releases were distributed and articles were published in the media.
-Didimo created content about humanizing technology and ethics for digital humans and disclosed it to position Didimo’s spokesperson as an opinion maker on this topic.
-Didimo has increased its social presence in the second year, increasing the number of followers from 1800 to more than 2700.