Photothermal effects in plasmonic nanostructures have great potentials in applications for photothermal cancer therapy, optical storage, thermo-photovoltaics, etc. However, the transient temperature behavior of a nanoscale material system during an ultrafast photothermal process has rarely been accurately investigated. Here a heat transfer model is constructed to investigate the temporal and spatial variation of temperature in plasmonic gold nanostructures. First, as a benchmark scenario, we study the light-induced heating of a gold nanosphere in water and calculate the relaxation time of the nanosphere excited by a modulated light. Second, we investigate heating and reshaping of gold nanoparticles in a more complex metamaterial absorber structure induced by a nanosecond pulsed light. The model shows that the temperature of the gold nanoparticles can be raised from room temperature to >795 K in just a few nanoseconds with a low light luminance, owing to enhanced light absorption through strong plasmonic resonance. Such quantitative predication of temperature change, which is otherwise formidable to measure experimentally, can serve as an excellent guideline for designing devices for ultrafast photothermal applications.

No abstract

Most work on temporal action detection is formulated as an offline problem, in which the start and end times of actions are determined after the entire video is fully observed. However, important real-time applications including surveillance and driver assistance systems require identifying actions as soon as each video frame arrives, based only on current and historical observations. In this paper, we propose a novel framework, Temporal Recurrent Network (TRN), to model greater temporal context of a video frame by simultaneously performing online action detection and anticipation of the immediate future. At each moment in time, our approach makes use of both accumulated historical evidence and predicted future information to better recognize the action that is currently occurring, and integrates both of these into a unified end-to-end architecture. We evaluate our approach on two popular online action detection datasets, HDD and TVSeries, as well as another widely used dataset, THUMOS'14. The results show that TRN significantly outperforms the state-of-the-art.

No abstract

Driving Scene understanding is a key ingredient for intelligent transportation systems. To achieve systems that can operate in a complex physical and social environment, they need to understand and learn how humans drive and interact with traffic scenes. We present the Honda Research Institute Driving Dataset (HDD), a challenging dataset to enable research on learning driver behavior in real-life environments. The dataset includes 104 hours of real human driving in the San Francisco Bay Area collected using an instrumented vehicle equipped with different sensors. We provide a detailed analysis of HDD with a comparison to other driving datasets. A novel annotation methodology is introduced to enable research on driver behavior understanding from untrimmed data sequences. As the first step, baseline algorithms for driver behavior detection are trained and tested to demonstrate the feasibility of the proposed task.

Bladder cancer is a common urologic cancer whose incidence continues to rise annually. Urinary microparticles are an attractive material for noninvasive bladder cancer biomarker discovery. In this study, we applied isotopic dimethylation labeling coupled with liquid chromatography-tandem mass spectrometry (LC-MS/MS) to discover bladder cancer biomarkers in urinary microparticles isolated from hernia (control) and bladder cancer patients. This approach identified 2964 proteins based on more than two distinct peptides, of which 2058 had not previously been reported as constituents of human urine exosomes/microparticles. A total of 107 differentially expressed proteins were identified as candidate biomarkers. Differences in the concentrations of 29 proteins (41 signature peptides) were precisely quantified by LC-MRM/MS in 48 urine samples of bladder cancer, hernia, and urinary tract infection/hematuria. Concentrations of 24 proteins changed significantly (p<0.05) between bladder cancer (n=28) and hernia (n=12), with area-under-the-curve values ranging from 0.702 to 0.896. Finally, we quantified tumor-associated calcium-signal transducer 2 (TACSTD2) in raw urine specimens (n=221) using a commercial ELISA and confirmed its potential value for diagnosis of bladder cancer. Our study reveals a strong association of TACSTD2 with bladder cancer and highlights the potential of human urinary microparticles in the noninvasive diagnosis of bladder cancer.

Radiation absorbers have increasingly been attracting attention as crucial components for controllable thermal emission, energy harvesting, modulators, etc. However, it is still challenging to realize thin absorbers which can operate over a wide spectrum range. Here, we propose and experimentally demonstrate thin, broadband, polarization-insensitive and omnidirectional absorbers working in the near-infrared range. We choose titanium (Ti) instead of the commonly used gold (Au) to construct nano-disk arrays on the top of a silicon dioxide (SiO2) coated Au substrate, with the quality (Q) factor of the localized surface plasmon (LSP) resonance being decreased due to the intrinsic high loss of Ti. The combination of this low-Q LSP resonance and the propagating surface plasmon (PSP) excitation resonance, which occur at different wavelengths, is the fundamental origin of the broadband absorption. The measured (at normal light incidence) absorption is over 90% in the wavelength range from 900 nm to 1825 nm, with high absorption persisting up to the incident angle of ~40°. The demonstrated thin-film absorber configuration is relatively easy to fabricate and can be realized with other properly selected materials.

A urine sample preparation workflow for the iTRAQ (isobaric tag for relative and absolute quantitation) technique was established. The reproducibility of this platform was evaluated and applied to discover proteins with differential levels between pooled urine samples from nontumor controls and three bladder cancer patient subgroups with different grades/stages (a total of 14 controls and 23 cancer cases in two multiplex iTRAQ runs). Combining the results of two independent clinical sample sets, a total of 638 urine proteins were identified. Among them, 55 proteins consistently showed >2-fold differences in both sample sets. Western blot analyses of individual urine samples confirmed that the levels of apolipoprotein A-I (APOA1), apolipoprotein A-II, heparin cofactor 2 precursor and peroxiredoxin-2 were significantly elevated in bladder cancer urine specimens (n = 25-74). Finally, we quantified APOA1 in a number of urine samples using a commercial ELISA and confirmed again its potential value for diagnosis (n = 126, 94.6% sensitivity and 92.0% specificity at a cutoff value of 11.16 ng/mL) and early detection (n = 71, 83.8% sensitivity and 94.0% specificity). Collectively, our results provide the first iTRAQ-based quantitative profile of bladder cancer urine proteins and represent a valuable resource for the discovery of bladder cancer markers.