Robust observation detection for single object tracking: deterministic and probabilistic patch-based approaches

Optimal Training Configurations of a CNN-LSTM-Based Tracker for a Fall Frame Detection System

In recent years, there has been an immense amount of research into fall event detection. Generally, a fall event is defined as a situation in which a person unintentionally drops down onto a lower surface. It is crucial to detect the occurrence of fall events as early as possible so that any severe fall consequences can be minimized. Nonetheless, a fall event is a sporadic incidence that occurs seldomly that is falsely detected due to a wide range of fall conditions and situations. Therefore, an automated fall frame detection system, which is referred to as the SmartConvFall is proposed to detect the exact fall frame in a video sequence. It is crucial to know the exact fall frame as it dictates the response time of the system to administer an early treatment to reduce the fall's negative consequences and related injuries. Henceforth, searching for the optimal training configurations is imperative to ensure the main goal of the SmartConvFall is achieved. The proposed SmartConvFall consists of two parts, which are object tracking and instantaneous fall frame detection modules that rely on deep learning representations. The first stage will track the object of interest using a fully convolutional neural network (CNN) tracker. Various training configurations such as optimizer, learning rate, mini-batch size, number of training samples, and region of interest are individually evaluated to determine the best configuration to produce the best tracker model. Meanwhile, the second module goal is to determine the exact instantaneous fall frame by modeling the continuous object trajectories using the Long Short-Term Memory (LSTM) network. Similarly, the LSTM model will undergo various training configurations that cover different types of features selection and the number of stacked layers. The exact instantaneous fall frame is determined using an assumption that a large movement difference with respect to the ground level along the vertical axis can be observed if a fall incident happened. The proposed SmartConvFall is a novel technique as most of the existing methods still relying on detection rather than the tracking module. The SmartConvFall outperforms the state-of-the-art trackers, namely TCNN and MDNET-N trackers, with the highest expected average overlap, robustness, and reliability metrics of 0.1619, 0.6323, and 0.7958, respectively. The SmartConvFall also managed to produce the lowest number of tracking failures with only 43 occasions. Moreover, a three-stack LSTM delivers the lowest mean error with approximately one second delay time in locating the exact instantaneous fall frame. Therefore, the proposed SmartConvFall has demonstrated its potential and suitability to be implemented for a real-time application that could help to avoid any crucial fall consequences such as death and internal bleeding if the early treatment can be administered.

Object tracking using adaptive covariance descriptor and clustering-based model updating for visual surveillance

We propose a novel approach for tracking an arbitrary object in video sequences for visual surveillance. The first contribution of this work is an automatic feature extraction method that is able to extract compact discriminative features from a feature pool before computing the region covariance descriptor. As the feature extraction method is adaptive to a specific object of interest, we refer to the region covariance descriptor computed using the extracted features as the adaptive covariance descriptor. The second contribution is to propose a weakly supervised method for updating the object appearance model during tracking. The method performs a mean-shift clustering procedure among the tracking result samples accumulated during a period of time and selects a group of reliable samples for updating the object appearance model. As such, the object appearance model is kept up-to-date and is prevented from contamination even in case of tracking mistakes. We conducted comparing experiments on real-world video sequences, which confirmed the effectiveness of the proposed approaches. The tracking system that integrates the adaptive covariance descriptor and the clustering-based model updating method accomplished stable object tracking on challenging video sequences.

Intelligent neonatal monitoring based on a virtual thermal sensor

BACKGROUND

Temperature measurement is a vital part of daily neonatal care. Accurate measurements are important for detecting deviations from normal values for both optimal incubator and radiant warmer functioning. The purpose of monitoring the temperature is to maintain the infant in a thermoneutral environmental zone. This physiological zone is defined as the narrow range of environmental temperatures in which the infant maintains a normal body temperature without increasing his or her metabolic rate and thus oxygen consumption. Although the temperature measurement gold standard is the skin electrode, infrared thermography (IRT) should be considered as an effortless and reliable tool for measuring and mapping human skin temperature distribution and assist in assessing thermoregulatory reflexes.

METHODS

Body surface temperature was recorded under several clinical conditions using an infrared thermography imaging technique. Temperature distributions were recorded as real-time video, which was analyzed to evaluate mean skin temperatures. Emissivity variations were considered for optimal neonatal IRT correction for which the compensation vector was overlaid on the tracking algorithm to improve the temperature reading. Finally, a tracking algorithm was designed for active follow-up of the defined region of interest over a neonate's geometry.

RESULTS

The outcomes obtained from the thermal virtual sensor demonstrate its ability to accurately track different geometric profiles and shapes over the external anatomy of a neonate. Only a small percentage of the motion detection attempts failed to fit tracking scenarios due to the lack of a properly matching matrix for the ROI profile over neonate's body surface.

CONCLUSIONS

This paper presents the design and implementation of a virtual temperature sensing application that can assist neonatologists in interpreting a neonate's skin temperature patterns. Regarding the surface temperature, the influence of different environmental conditions inside the incubator has been confirming.

Incremental structured dictionary learning for video sensor-based object tracking

To tackle robust object tracking for video sensor-based applications, an online discriminative algorithm based on incremental discriminative structured dictionary learning (IDSDL-VT) is presented. In our framework, a discriminative dictionary combining both positive, negative and trivial patches is designed to sparsely represent the overlapped target patches. Then, a local update (LU) strategy is proposed for sparse coefficient learning. To formulate the training and classification process, a multiple linear classifier group based on a K-combined voting (KCV) function is proposed. As the dictionary evolves, the models are also trained to timely adapt the target appearance variation. Qualitative and quantitative evaluations on challenging image sequences compared with state-of-the-art algorithms demonstrate that the proposed tracking algorithm achieves a more favorable performance. We also illustrate its relay application in visual sensor networks.