Robot navigation in crowded public spaces is a complex task that requires addressing variety of engineering and human factors challenges. These challenges have motivated great amount research resulting important developments for the fields robotics human-robot interaction over past three decades. Despite significant progress massive recent interest, we observe number remaining prohibit seamless deployment autonomous robots environments. In this survey article, organize existing into set...

We propose a Convolutional Neural Network-based approach to learn, detect,and extract patterns in sequential trajectory data, known here as Social Pattern Extraction Convolution (Social-PEC). A set of experiments carried out on the human prediction problem shows that our model performs comparably state art and outperforms some cases. More importantly,the proposed unveils obscurity previous use pooling layer, presenting way intuitively explain decision-making process.

Obstacle avoidance is one of the essential and indispensable functions for autonomous mobile robots. Most existing solutions are typically based on single condition constraint cannot incorporate sensor data in a real-time manner, which often fail to respond unexpected moving obstacles dynamic unknown environments. In this paper, novel multi-constraints obstacle method using Light Detection Ranging(LiDAR) proposed, able to, latest estimation robot pose environment, find sub-goal defined by...

The last few years have witnessed the great success of generative adversarial networks (GANs) in synthesizing high-quality photorealistic face images. Many recent 3D facial texture reconstruction works often pursue higher resolutions and ignore occlusion. We study problem detailed under occluded scenes. This is a challenging problem; currently, collection such large scale high resolution dataset still very costly. In this work, we propose deep learning based approach for that does not...

The huge advantage of in-pipe robots is that they are able to measure from inside the pipes, and sense geometry, appearance radiometry directly. downside inability know precise, absolute position measurements in very long pipe runs. This paper develops unprecedented localization required for this purpose.

Over the past few years, single-view 3D face reconstruction methods can produce beautiful models. Nevertheless,the input of these works is unobstructed faces.We describe a system designed to reconstruct convincing texture in case occlusion.Motivated by parsing facial features,we propose complete map generation method guided landmarks.We estimate 2D structure reasonable position occlusion area,which used for construction texture.An excellent anti-occlusion should ensure authenticity...

3D face reconstruction technology aims to generate a stereo model naturally and realistically. Previous deep approaches are typically designed convincing textures cannot generalize well multiple occluded scenarios simultaneously. By introducing bump mapping, we successfully added mid-level details coarse faces. More innovatively, our method takes into account occlusion scenarios. Thus on top of common approaches, in this paper propose unified framework handle types obstruction simultaneously...

Single-view 3D face reconstruction is a fundamental Computer Vision problem of extraordinary difficulty. Current systems often assume the input unobstructed faces which makes their method not suitable for in-the-wild conditions. We present performing that removes eyeglasses from single image. Existing facial methods fail to remove automatically generating photo-realistic "in-the-wild".The innovation our lies in process identifying area robustly and it intelligently. In this work, we estimate...

Obstacle avoidance is one of the essential and indispensable functions for autonomous mobile robots. Most existing solutions are typically based on single condition constraint cannot incorporate sensor data in a real-time manner, which often fail to respond unexpected moving obstacles dynamic unknown environments. In this paper, novel multi-constraints obstacle method using Light Detection Ranging(LiDAR) proposed, able to, latest estimation robot pose environment, find sub-goal defined by...

As more and robots are envisioned to cooperate with humans sharing the same space, it is desired for be able predict others' trajectories navigate in a safe self-explanatory way. We propose Convolutional Neural Network-based approach learn, detect, extract patterns sequential trajectory data, known here as Social Pattern Extraction Convolution (Social-PEC). A set of experiments carried out on human prediction problem shows that our model performs comparably state art outperforms some cases....