The FaceDirector system enables directors to fine-tune performances in post-production, saving time and money by avoiding reshoots. It combines facial expressions and audio cues for optimal synchronization, allowing users to generate novel versions of performances.
Researchers developed a new technique called photogeometric scene flow (PGSF) that combines three computer vision methods to capture high-quality and detailed facial features. The method produces superior results in capturing facial details, making it extremely valuable for realistic facial reconstructions.
A new visual authoring tool is being developed to enable both experts and novices to create digital story worlds. The tool guides users through three main steps: story world creation, character and object authoring, and event creation.
A new technology developed by Carnegie Mellon University and Disney Research enables smartwatches to automatically recognize what objects users are touching. The technique, called EM-Sense, takes advantage of the body's natural electrical conductivity to detect and identify objects, creating opportunities for context-aware apps.
Disney researchers developed mobile apps that add computer-generated animation, sound, and visuals to traditional activities like coloring books, scavenger hunts, and team games. The apps aim to create a seamless fusion of the real and virtual worlds, promoting creative play and exploration.
Disney Research develops tactile animation, an animation software-like tool to create rich haptic sensations. The new process allows designers to focus on unfolding haptic effects over space and time, reducing the need for individually controlling actuators.
A new system developed by Disney Research and Carnegie Mellon University enables users to design customized walking robots using intuitive editing tools. The system ensures the robot moves as intended, allowing users to adjust its gait and behavior.
The tool uses a library of base shapes to enable users to think creatively without tedious trial and error. Researchers used the tool to design and build various shapes, including a turtle with 123 pieces, which was assembled in about 2.5 hours.
Researchers at Disney Research have developed an automated method called AutoConnect that can design custom connectors for 3D-printable objects. The tool allows users to input the dimensions and weight of two objects, as well as how they should be aligned when connected.
Researchers at Disney Animation Studios developed a multigrid method that speeds up cloth simulation six to eight times faster than conventional methods. The technique enables more realistic look and behavior of cloth in animations, making it suitable for simulations requiring lots of detail like clothing worn by characters.
The Disney Research app uses augmented reality to turn coloring books into a 3D experience, increasing motivation and connection in children. The app accurately colors the 3D character based on the user's coloring, creating an illusion of real-time application.
The soft skin modules use air-filled cavities to cushion collisions and provide pressure feedback for grasping delicate objects. The researchers successfully tested the modules on a disposable plastic cup, a roll of printer paper, and a piece of tofu, reducing peak forces by 32-52% and side impacts by 26-37%.
The new algorithm enables filmmakers to determine the optimal number and location of cameras for capturing a scene, amassing data that can be used during post-production. It allows for virtual relighting, generating new viewpoints, and digitally recovering detailed 3D shapes and layouts.
Researchers at Disney Research have developed a networking technology that enables LED lights to communicate with each other and the Internet using visible light signals. The technology has the potential to create smart environments and enable IoT applications, such as remote monitoring of objects.
Scientists have developed a network of energy-harvesting sensor nodes equipped with onboard cameras that can automatically determine each camera's pose and location. This capability enables large-scale sensor networks to operate without batteries or external power, making them ideal for the Internet of Things (IoT) applications such as...
The new method uses a hybrid approach employing three different rendering techniques depending on the scale at which the object is viewed. It allows for accurate rendering while reducing computation time, speeding up processing by tens or hundreds of times.
Researchers at Disney Research have developed a method to create 3D-printed objects with varying levels of elasticity, enabling the creation of deformable toys and soft robots. By controlling the small-scale structure of the material, they can produce complex microstructures that mimic the properties of metamaterials.
Researchers at Disney Research developed an efficient method to collect and filter video pixels from multiple frames, enabling deblurring, denoising, and object removal. This technique uses scene-space processing to compensate for 3-D inaccuracies in real-world videos.
LinkEdit enables intuitive and predictable changes to planar linkage shape and motion, preserving critical features like walking gait. The software uses symbolic kinematics for efficient computation, allowing users to modify links, joints, and linkage size.
Researchers developed a tool enabling art directors to control computer programs using verbal descriptors like 'silky' and 'wrinkly'. The system improved the process of creating garments with desired properties, reducing laborious tweaking of technical parameters.
The OmniAD technique uses video recording data to reconstruct the 3D motion of an object and infer aerodynamic forces. This allows for realistic simulation of light, 3-D objects in videogames and animations.
Disney Research has developed a new adaptive rendering method that efficiently corrects for erroneous pixels while preserving crisp detail, resulting in faster production times. The approach uses sparse modeling to predict additional pixels based on patterns identified in a small number of image pixels.
Scientists capture subtle eyelid motions and create plausible folding and stretching in areas not directly captured during data collection. The method improves photorealistic actors for films by accurately determining the location of wrinkles and predicting their change based on eye region history.
A new performance capture method developed at Disney Research captures facial features in high resolution, including wrinkles, and operates in real-time with a single video camera. This allows for realistic virtual characters and improved emotional expression conveyance.
Researchers at Disney Research Zurich developed a method for accurate video object segmentation using a click-and-drag interface that enables human editors to work efficiently with state-of-the-art algorithms. This approach achieves higher accuracies than fully automated systems by keeping a human in the loop.
A research group at Disney Research Pittsburgh developed a computer vision system that continuously improves its ability to recognize objects by picking up hints from videos. The system outperformed other methods in detecting various objects, including microwave ovens and stoves.
A new method developed by Disney Research Pittsburgh and Boston University can recognize actions in videos more accurately than previous methods. The algorithm represents actions as space-time patterns, allowing it to account for variations in execution.
Researchers at Disney Research Pittsburgh have developed a 2-legged robot that can mimic an animated character's walking motion, using a combination of 3D-printed links and servo motors. The robot is designed to duplicate the character's gait as closely as possible, while also being stable and physically realizable.
The Disney Research algorithm corrects for parallax and image warping, eliminating ghosting and other distortions in video panoramas. It generates high-quality panoramic videos with resolutions up to 100 megapixels using unstructured camera arrays.
The study developed a method enabling dialogue replacement for automated video redubbing, utilizing facial movements associated with speech sounds. The approach produced far more plausible alternative word sequences than conventional methods.
A team from Disney Research and Carnegie Mellon University have devised a 3D printer that layers together laser-cut sheets of fabric to form soft, squeezable objects with complex geometries and integrated circuitry. The printer combines fabrics and wiring to create interactive objects such as bunnies, doll clothing, and phone cases.
Researchers at Disney have developed a system that uses battery-free RFID tags to detect how people interact with everyday objects. The IDSense system can track multiple objects simultaneously and infer movement patterns with high accuracy, opening up new possibilities for interactive play, smart homes, and consumer behavior analysis.
Disney Research develops a new platform to automate aspects of authoring interactive narratives, allowing users more control over storylines and resolving narrative inconsistencies. The IBTs enable authors to create free-form experiences with multiple story arcs.
Disney Research has developed a robotic camera system that can learn from human operators to better frame shots of a basketball game. The system uses machine learning algorithms to recognize the relationship between player locations and corresponding camera configurations.
A new automated method developed by Disney Research uses AI to select and order photos in a way that makes narrative sense, telling a compelling story. The system learns principles of selecting and ordering photos from large collections and can customize the process for individual preferences.
Disney Research creates a method for automatically reviewing nighttime giraffe behavior videos and creating a summary of unusual behaviors. The system can detect rare events in real-time, allowing wildlife experts to better understand giraffe health and safety.
Researchers used machine learning approaches to build models that can make accurate predictions of player behavior, including passes and shots. The models also identified team formations and could accurately detect and visualize them, enabling teams to be identified based on their style of play.
Scientists at Disney Research and Université Laval have developed a method to automatically estimate illumination conditions depicted in a collection of photographs. This enables the realistic insertion of 3D objects into images, simplifying photo editing tasks.
A new image processing technique developed by Disney Research Zurich enables video effects like detailed actor faces despite lighting changes. The local tone mapping method preserves rich visual detail while eliminating ghosting and camera noise.
Scientists at Disney Research Zurich developed a technique to capture detailed eye features, including shape, texture, and iris deformation. This enables the creation of realistic digital eyes that can respond to lighting changes.
The Disney Research library contains over 50 'feel effects,' including rain, a walking cat, and a beating heart. Novice users can create and share their own haptic effects using the FeelCraft software.
The new transmission enables robot arms to be light and strong while maintaining precise control. It also allows for 'give' in the limbs, facilitating soft interactions between people and robots.
A Disney Research team created an algorithm that automatically edits footage from multiple cameras, following established rules of cinematography and identifying the most interesting content. The resulting videos capture essential action and are often similar in spirit to those produced by professionals.
Researchers at Disney Research Zurich and the University of Zaragoza have developed a method to capture an individual's hairstyle in 3D-printed figurines, improving realism and individuality. The system uses color images to compute coarse geometry and stylize hair while preserving defining features.
Researchers have developed a new tool to synchronize multiple video clips based on the visual content of the videos. This makes it easier to align clips without relying on timecodes or external markers. The method can be used for various applications, including movie production and visual effects.
Researchers at Disney Research Zurich have developed an algorithm that can optimize the mass distribution within objects to enable a stable spin around a desired axis, applicable to various shapes and materials. The approach has potential applications beyond customized designs for spinning toys.
Researchers at Disney Research Pittsburgh established a library of 40 feel effects matched to descriptions that designers can understand. The study aimed to provide a common vocabulary for designing haptic feedback experiences.
Researchers developed tools for designing mechanisms that achieve desired cyclical motion in planar mechanical characters and linkage-based sculptures. The systems enable casual users to create animated characters using 3D printing techniques, making it possible to bring creations to life.
The new interactive tool allows designers to create intricate shapes by sketching the structure's shape and seam placement, generating patterns for flat panels that assume the desired shape when inflated. The system is validated with seven varied shapes, including an elephant and a flower, and demonstrates efficient design capabilities.
Researchers at Disney Research have created electrostatic loudspeakers that can be printed in any shape using 3D technology, enabling objects to produce sound and even ultrasound. The technology has potential applications in interactive systems, toys, and games.