Search Results

4D Spontaneous Facial Expression Database Background: Because posed and un-posed (aka “spontaneous”) 3D facial expressions differ along several dimensions including complexity and timing, well-annotated 3D video of un-posed facial behavior is needed. The BP4D dataset comprises videos of spontaneous facial expressions in a diverse...

Patented radio receiver and a method to compensate for distortion Technology Overview: The patented technology leverages artificial intelligence, specifically deep neural networks (DNNs) at the receiver to equalize signals transmitted through communication channels dominated by nonlinear distortion (NLD). DNN may enable superior receiver equalization...

Novel algorithm uses biological signals to determine the authenticity of videos Background: In recent years, technology development of deep learning has advanced to the level of high realism in creating synthetic videos. Such deep fake videos pose a big threat to video authentication for security, safety, privacy protection, as well as health-care...

This technology allows servers to be updated without downtime, solving the common problem that updating hypervisors or operating systems usually requires reboots or temporarily stopping applications. Instead of copying large amounts of memory, it switches memory control in place, enabling updates in under a second while applications keep running. Background:...

This technology enables humans to negotiate robot actions instantly through an intuitive augmented reality interface. By visualizing robot intentions and allowing direct constraint input, it removes ambiguity and accelerates collaboration. The result is faster, clearer, and more flexible human-robot teamwork in dynamic, real-world environments. Background:...

This invention introduces the MMSE HR dataset along with a Self Adaptive Matrix Completion algorithm for remote heart rate estimation from facial videos in natural conditions. By combining synchronized spontaneous expression data with physiological signals, it enables accurate, non contact heart rate measurement and real time emotion analysis for robust...

An autonomous system to learn how to perform a complex task from past experiences through reinforcement learning without human intervention Background: Recent advances in autonomy technology have promoted the widespread emergence of autonomous systems in various domains such as domestic robots, self-driving vehicles, and financial management...

Semisupervised Autoencoder for Sentiment Analysis that reduces bias and requires minimal training Background: They act as the feature learning methods by reconstructing inputs with respect to a given loss function. In a neural network implementation of autoencoders, the hidden layer is taken as the learned feature. While it is often trivial to...