Search Results for « Temporal Smoothing »

In order to realize the mapping from spatial information to temporal information, Temporal Smoothing (TS) Learning is proposed. In this learning, the output of a learning unit, to which sensory signals are given as input, is trained to be smooth along time. In other words, the learning unit is trained so as that the second time derivative of the output itself becomes 0.This learning can be applied to integrate local sensory signals into an analog spatial signal. It also can be used that an agent learn evaluation function in delayed reinforcement learning on behalf of TD Learning(8)(7) when only one target state is chosen. When a neural network was employed as a learning unit and visual signals were given as inputs directly, the hidden neurons in the neural network represented spatial information and had a adaptability of changing the representation according to the agent's motion characteristics.

Multi-task learning models based on temporal smoothness assumption, in which each time point of a sequence of time points concerns a task of prediction, assume the adjacent tasks are similar to each other. However, the effect of outliers is not taken into account. In this paper, we show that even only one outlier task will destroy the performance of the entire model. To solve this problem, we propose two Robust Temporal Smoothness (RoTS) frameworks. Compared with the existing models based on temporal relation, our methods not only chase the temporal smoothness information but identify outlier tasks, however, without increasing the computational complexity. Detailed theoretical analyses are presented to evaluate the performance of our methods. Experimental results on synthetic and real-life datasets demonstrate the effectiveness of our frameworks. We also discuss several potential specific applications and extensions of our RoTS frameworks.

The automatization of weather warnings is an ongoing endeavor and plays an important role in the program RainBoW ("Risk-based, Application-oriented and INdividualizaBle Provision of Optimized Warning Information"), which aims to renew the warning system of the German Meteorological Service (Deutscher Wetterdienst, DWD). Automated warnings are often generated at relatively high frequency, depending on the availability of updated forecast data. For a fixed forecast date, these updates of the warning information may lead to frequent changes of, for example, the warning level, that could confuse end-users. One method to circumvent frequent jumps in the warning level is called smoothing. This work presents a prototype system for temporal smoothing of automated weather warnings, focusing on wind gusts and thunderstorm events to gain temporal stability and to minimize fluctuating warnings.For wind gusts, this study uses automatically generated warnings based on ensemble wind gust forecasts from the in-house ICON model, which is updated hourly. Smoothing of thunderstorm warnings is based on the output of the NowCastMIX nowcasting application, which predicts the development of thunderstorm cells and their characteristics in 5-minute increments for up to one hour. The techniques to enhance the temporal warning stability are applied to a grid-based framework. For thunderstorm and wind gust warnings, temporal smoothing is achieved by evaluating the maximum warning level across a series of the most recent forecast data for the same forecast date. Wind-related parameters that complement the warning, including gust speeds and wind direction, undergo a weighted smoothing procedure that accounts for the higher precision of more recent forecasts. This methodology aims to reduce unnecessary fluctuations in warning outputs without compromising the responsiveness of the system to genuine changes in weather conditions.This prototype for smoothing wind gust warnings contributes to the ongoing effort to automatize weather warnings at the DWD and is an important step towards more user-friendly warning-products.

Sparse representation has been popular in visual tracking recently for its robustness and accuracy. However, for most conventional sparse coding based trackers, the target candidates are considered independently between consecutive frames. This paper shows that the temporal correlation of these frames can be exploited to improve the performance of tracking and makes the tracker more robust to noise. Furthermore, to improve the tracking speed, we revisit a more efficient method for ℓ1norm problem, marginal regression, which can solve the sparse coding problem more efficiently. Consequently we can realize real-time tracking based on the temporal smooth sparse representation. Extensive experiments have been done to demonstrate the effectiveness and efficiency of our method.

We propose an automatic virtual cinematography method that takes a continuous optimization approach. A suitable camera pose or path is determined automatically by computing the minima of an objective function to obtain some desired parameters, such as those common in live action photography or cinematography. Multiple objective functions can be combined into a single optimizable function, which can be extended to model the smoothness of the optimal camera path using an active contour model. Our virtual cinematography technique can be used to find camera paths in either scripted or unscripted scenes, both with and without smoothing, at a relatively low computational cost.

This paper proposes a method for machine learning from unlabeled data in the form of a time-series. The mapping that is learned is shown to extract slowly evolving information that would be useful for control applications, while efficiently filtering out unwanted, higher-frequency noise. The method consists of training a feedforward artificial neural network with backpropagation using two opposing objectives. The first of these is to minimize the squared changes in activations between time steps of each unit in the network. This "temporal smoothing" has the effect of correlating inputs that occur close in time with outputs that are close in the L2-norm. The second objective is to maximize the log determinant of the covariance matrix of activations in each layer of the network. This objective ensures that information from each layer is passed through to the next. This second objective acts as a balance to the first, which on its own would result in a network with all input weights equal to zero.

We present a new approach to multi-signal gesture recognition that attends to simultaneous body and hand movements. The system examines temporal sequences of dual-channel input signals obtained via statistical inference that indicate 3D body pose and hand pose. Learning gesture patterns from these signals can be quite challenging due to the existence of long-range temporal-dependencies and low signal-to-noise ratio (SNR). We incorporate a Gaussian temporal-smoothing kernel into the inference framework, capturing long-range temporal-dependencies and increasing the SNR efficiently. An extensive set of experiments was performed, allowing us to (1) show that combining body and hand signals significantly improves the recognition accuracy; (2) report on which features of body and hands are most informative; and (3) show that using a Gaussian temporal-smoothing significantly improves gesture recognition accuracy.