Existing state-of-the-art salient object detection networks rely on aggregating multi-level features of pretrained convolutional neural networks (CNNs). Compared to high-level features, low-level features contribute less to performance but cost more computations because of their larger spatial resolutions. In this paper, we propose a novel Cascaded Partial Decoder (CPD) framework for fast and accurate salient object detection. On the one hand, the framework constructs partial decoder which discards larger resolution features of shallower layers for acceleration. On the other hand, we observe that integrating features of deeper layers obtain relatively precise saliency map. Therefore we directly utilize generated saliency map to refine the features of backbone network. This strategy efficiently suppresses distractors in the features and significantly improves their representation ability. Experiments conducted on five benchmark datasets exhibit that the proposed model not only achieves state-of-the-art performance but also runs much faster than existing models. Besides, the proposed framework is further applied to improve existing multi-level feature aggregation models and significantly improve their efficiency and accuracy.

No abstract

The present study examined the psychometric properties of scores from a direct measure of behavioral regulation, the Head-Toes-Knees-Shoulders task (HTKS) with 3- to 6-year-old children in the U.S., Taiwan, South Korea, and China. Specifically, we investigated (1) the nature and variability of HTKS scores including relations to teacher-rated classroom behavioral regulation, and (2) relations between the HTKS and early mathematics, vocabulary, and literacy skills. Higher HTKS scores were significantly related to higher teacher ratings of classroom behavioral regulation in the U.S. and South Korea but not in Taiwan and China. Also, higher HTKS scores were significantly related to higher early mathematics, vocabulary, and literacy skills beyond the influence of demographic variables and teacher-rated classroom behavioral regulation. These initial findings suggest that HTKS scores may be interpreted as reflecting early behavioral regulation in these four societies, and that behavioral regulation is important for early academic success in the U.S. and in Asian countries.

One of the critical challenges of object counting is the dramatic scale variations, which is introduced by arbitrary perspectives. We propose a reverse perspective network to solve the scale variations of input images, instead of generating perspective maps to smooth final outputs. The reverse perspective network explicitly evaluates the perspective distortions, and efficiently corrects the distortions by uniformly warping the input images. Then the proposed network delivers images with similar instance scales to the regressor. Thus the regression network doesn't need multiscale receptive fields to match the various scales. Besides, to further solve the scale problem of more congested areas, we enhance the corresponding regions of ground-truth with the evaluation errors. Then we force the regressor to learn from the augmented ground-truth via an adversarial process. Furthermore, to verify the proposed model, we collected a vehicle counting dataset based on Unmanned Aerial Vehicles (UAVs). The proposed dataset has fierce scale variations. Extensive experimental results on four benchmark datasets show the improvements of our method against the state-of-the-arts.

In crowd counting datasets, the location labels are costly, yet, they are not taken into the evaluation metrics. Besides, existing multi-task approaches employ high-level tasks to improve counting accuracy. This research tendency increases the demand for more annotations. In this paper, we propose a weakly-supervised counting network, which directly regresses the crowd numbers without the location supervision. Moreover, we train the network to count by exploiting the relationship among the images. We propose a soft-label sorting network along with the counting network, which sorts the given images by their crowd numbers. The sorting network drives the shared backbone CNN model to obtain density-sensitive ability explicitly. Therefore, the proposed method improves the counting accuracy by utilizing the information hidden in crowd numbers, rather than learning from extra labels, such as locations and perspectives. We evaluate our proposed method on three crowd counting datasets, and the performance of our method plays favorably against the fully supervised state-of-the-art approaches.