Sorry, you need to enable JavaScript to visit this website.

Data reduction algorithms, including matrix factorization techniques, represent an essential component of many ML systems.One popular paradigm of matrix factorizations includes Non-Negative Matrix Factorization (NMF).

In this paper, we propose an efficient convolution neural network to enhance the quality of video compressed by HEVC standard. The model is composed of a max-pooling module and a hierarchical feature fusion module. The max-pooling module extracts feature from different scales and enlarges the receptive field of the model without stacking too many convolution layers. And the hierarchical feature fusion module accurately aligns features from different scales and fuses them efficiently.

We propose a data-driven secure wireless communication scheme, in which the goal is to transmit a signal to a legitimate receiver with minimal distortion, while keeping some information about the signal private from an eavesdropping adversary. When the data distribution is known, the optimal trade-off between the reconstruction quality at the legitimate receiver and the leakage to the adversary can be characterised in the information theoretic asymptotic limit.

Time-of-Flight (ToF) cameras provide a fast and robust way of acquiring the 3D shape of real scenes. Dense depth images can be generated at tens of frame per second. 3D shapes can be then segmented and objects classified, but can we directly sense the objects’ material using just a ToF camera? This live demonstration proves the answer to be affirmative. This possibility has only very recently been unveiled and we are, to the best of our knowledge, the first providing a live demonstrator showing the feasibility of this approach.

We present a compact data structure to represent both the duration and length of homogeneous segments of trajectories from moving objects in a way that, as a data warehouse, it allows us to efficiently answer cumulative queries. The division of trajectories into relevant segments has been studied in the literature under the topic of Trajectory Segmentation. In this paper, we design a data structure to compactly represent them and the algorithms to answer the more relevant queries.

This paper presents a novel deep architecture for weakly-supervised temporal action localization that predicts temporal boundaries with graph regularization. Our model not only generates segment-level action responses but also propagates segment-level responses to
neighborhood in a form of graph Laplacian regularization. Specifically, our approach consists of two sub-modules; a class activation
module to estimate the action score map over time through the action classifiers, and a graph regularization module to refine the