Search

In recent years, deep learning has made dramatic advances in computer vision field, especially in improving the performance of object detection as well as instance semantic segmentation. Still, multi-object tracking (MOT) remains a very challenging issue. Even in state-of-the-art deep learning-based object detectors, a preferred paradigm for MOT: tracking-by-detection, can only slightly improve the tracking performance. Pixel-level information is considered more precise and useful for tracking performance improvement than using conventional information, such as foreground or background content in a bounding box. However, the performance of current state-of-the-art models for automatically annotating pixel-level information is still far from the expectation of human beings. Therefore, we shall explore how multi-object tracking and segmentation (MOTS) is affected when the information obtained after applying instance semantic segmentation is incomplete. We propose a mask-guided two-streamed augmentation learning (MGTSAL) algorithm, which can be applied to TrackR-CNN to alleviate significant drop of MOTS performance when encountering incompletely segmented information. We evaluate the proposed approach on MOTS KITTI dataset, and our approach outperforms the baseline model TrackR-CNN in all our experimental settings. The promising experimental results and ablation study validate the effectiveness of the proposed approach.

The systolic time intervals of hearts are related to health. People with myocardial dysfunction will have a longer pre-ejection period (PEP) and a shorter left ventricle ejection time (LVET) than healthy people. The purpose of this paper is intended to detect PEP and LVET accurately from electrocardiography (ECG) and phonocardiogram (PCG). Generally, there are several kinds of noises from environment or breathing in PCG. It is necessary but difficult to extract the best signals we want. Our approach is to use a simple DSP-based method in PCG to detect aortic valve opening (AVO) and aortic valve closure (AVC) times as well as R-peak time in ECG. Then, the PEP and LVET can be calculated. We evaluated PEP and LVET of 72 files from 46 people. To the annotated data, the PEP range is around 5 ms to 100 ms; the LVET range is around 170 ms to 380 ms. Our PEP results have 63.96% accuracy within 20 ms absolute error and 91.74% accuracy within 40 ms absolute error; LVET have 75.48% accuracy within 40 ms absolute error and 93.53% accuracy within 80 ms absolute error.

For multi-target multi-camera recognition tasks, tracking of objects of interest is one of the essential yet challenging issues due to the fact that the task requires re-identifying identical targets across distinct views. Multi-target multi-camera tracking (MTMCT) applications span a wide range of variety (e.g. crowd behavior analysis, anomaly individual tracking and sport player tracking), so how to make the system perform real-time tracking becomes a crucial research issue. In this paper, we propose an online hierarchical algorithm for extreme clustering based MTMCT framework. The system can automatically create a dynamic gallery with real-time fashion by collecting appearance information of multi-object tracking in single-camera view. We evaluate the effectiveness and efficiency of our framework, and compare the state-of-the-art methods on MOT16 as well as DukeMTMC for single and multiple camera tracking. The high-frame-rate performance and promising tracking results confirm our system can be used in realworld applications.

Translating video to language is very challenging due to diversified video contents originated from multiple activities and complicated integration of spatio-temporal information. There are two urgent issues associated with the video-to-language translation problem. First, how to transfer knowledge learned from a more general dataset to a specific application domain dataset? Second, how to generate stable video captioning (or description) results under different sampling rates? In this paper, we propose a novel temporal embedding method to better retain temporal representation under different video sampling rates. We present a transfer learning method that combines a stacked LSTM encoder-decoder structure and a temporal embedding learning with soft-attention (TELSA) mechanism. We evaluate the proposed approach on two public datasets, including MSR-VTT and MSVD. The promising experimental results confirm the effectiveness of the proposed approach.

A series of Cyclo-olefin copolymer (COC) blends were prepared via solution blending process and showed various miscibility because of different norbornene content as well as copolymer microstructure. Two kinds of nano-silica particles, including hydrophilic silica (Hp-SiO2) and hydrophobic silica (Hb-SiO2) were inducted into COC blends to deform the phase separation domains of COC blends. The properties and morphologies of the composites were characterized after preparation. The tensile moduli of 8007 blended films increase with increasing 5013 or 6013 content. The tensile moduli of the COCs composites slightly decrease either with 1 wt% Np- or Nb-SiO2. But the tensile strength and elongation increase as the 6013 content higher than 30 wt% in the 8007/6013 composites with 1 wt% Nb-SiO2 particles. By adding either Hp- or Hb-SiO2 particles, the haze value decreased in the immiscible blends (8007/5013) but slightly increase in the miscible blends (8007/6013). For all COCs and their blends, the transmittances are all higher than 85% with 1 wt% Nb-SiO2 loading. The phase separated domains deformed with nanosilica in 8007/5013 composites which might be a reason for lower haze. In addition, transparent composite materials can be manufactured from immiscible COC blends containing nanoparticles.