Informative Data Selection With Uncertainty for Multimodal Object Detection.

Noise has always been nonnegligible trouble in object detection by creating confusion in model reasoning, thereby reducing the informativeness of the data. It can lead to inaccurate recognition due to the shift in the observed pattern, that requires a robust generalization of the models. To implement a general vision model, we need to develop deep learning models that can adaptively select valid information from multimodal data. This is mainly based on two reasons. Multimodal learning can break through the inherent defects of single-modal data, and adaptive information selection can reduce chaos in multimodal data. To tackle this problem, we propose a universal uncertainty-aware multimodal fusion model. It adopts a multipipeline loosely coupled architecture to combine the features and results from point clouds and images. To quantify the correlation in multimodal information, we model the uncertainty, as the inverse of data information, in different modalities and embed it in the bounding box generation. In this way, our model reduces the randomness in fusion and generates reliable output. Moreover, we conducted a completed investigation on the KITTI 2-D object detection dataset and its derived dirty data. Our fusion model is proven to resist severe noise interference like Gaussian, motion blur, and frost, with only slight degradation. The experiment results demonstrate the benefits of our adaptive fusion. Our analysis on the robustness of multimodal fusion will provide further insights for future research.

[A case-crossover study on air pollutants and the mortality of stroke].

OBJECTIVE: To investigate the short-term effect of particulate matter in air on the mortality of stroke. METHODS: Using time-stratified case-crossover study design, an association was examined between stroke mortality and particulate matter with aerodynamic diameter of < 10 microm (PM10) of 2002 - 2004 in Hangzhou city. Meanwhile, the acute health effect of other gaseous pollutants (sulfur dioxide, SO2 and nitrogen dioxide, NO2) was also analyzed. RESULTS: A total of 9906 deaths of stroke were included. The crude stroke mortality was 83.54 per 100 000. After being adjusted for meteorological factors, when an increase of 10 microg/m3 in PM10, SO2 and NO2 in three days was noticed, it appeared that the increases of mortality of stroke were 0.56% (95% CI: 0.14%-0.99%), 1.62% (95% CI: 0.26% - 3.01%) and 2.07% (95% CI: 0.54% - 3.62%) respectively. There was no distinct association in multi-pollutant models. In sensitivity analysis, the associations were found in all single-pollutant models but not statistically significant in multi-pollutant models after replacing the missing values. CONCLUSION: It is suggested that the short-term elevation in PM10 as well as SO2 and NO2 daily concentrations were related to the increase of stroke mortality in Hangzhou city.