The loss function, also known as cost function, is used for training a neural network or other
machine learning models. Over the past decade, researchers have designed many loss …

We motivate and present feature selective anchor-free (FSAF) module, a simple and
effective building block for single-shot object detectors. It can be plugged into single-shot …

Large-scale object detection datasets (eg, MS-COCO) try to define the ground truth
bounding boxes as clear as possible. However, we observe that ambiguities are still …

Non-maximum suppression (NMS) is essential for stateof-the-art object detectors to localize
object from a set of candidate locations. However, accurate candidate location sometimes is …

Instance segmentation, which has been required in various applications in recent years, is
aimed at reliable bounding box (bbox) detection (ie, localization) and stable mask prediction …

An urban smart lighting architecture allowing municipalities to manage and control public
street lighting lamps is here considered. The system is capable to autonomously modulate …

We propose a collection of three shift-based primitives for building efficient compact CNN-
based networks. These three primitives (channel shift, address shift, shortcut shift) can …

In the majority of object detection frameworks, the confidence of instance classification is
used as the quality criterion of predicted bounding boxes, like the confidence-based ranking …

We propose a collection of three shift-based primitives for building efficient compact CNN-
based networks. These three primitives (channel shift, address shift, shortcut shift) can …

Motivated by real-world deployment of drones for conservation, this paper advances the
state-of-the-art in security games with signaling. The well-known defender-attacker security …