Goto

Collaborating Authors

Shape Evasion: Preventing Body Shape Inference of Multi-Stage Approaches

arXiv.org Artificial Intelligence

Modern approaches to pose and body shape estimation have recently achieved strong performance even under challenging real-world conditions. Even from a single image of a clothed person, a realistic looking body shape can be inferred that captures a users' weight group and body shape type well. This opens up a whole spectrum of applications -- in particular in fashion -- where virtual try-on and recommendation systems can make use of these new and automatized cues. However, a realistic depiction of the undressed body is regarded highly private and therefore might not be consented by most people. Hence, we ask if the automatic extraction of such information can be effectively evaded. While adversarial perturbations have been shown to be effective for manipulating the output of machine learning models -- in particular, end-to-end deep learning approaches -- state of the art shape estimation methods are composed of multiple stages. We perform the first investigation of different strategies that can be used to effectively manipulate the automatic shape estimation while preserving the overall appearance of the original image.


gscnn

#artificialintelligence

Current state-of-the-art methods for image segmentation form a dense image representation where the color, shape and texture information are all processed together inside a deep CNN. This however may not be ideal as they contain very different type of information relevant for recognition. We propose a new architecture that adds a shape stream to the classical CNN architecture. The two streams process the image in parallel, and their information gets fused in the very top layers. Key to this architecture is a new type of gates that connect the intermediate layers of the two streams.


Combined discriminative and generative articulated pose and non-rigid shape estimation

Neural Information Processing Systems

Estimation of three-dimensional articulated human pose and motion from images is a central problem in computer vision. Much of the previous work has been limited by the use of crude generative models of humans represented as articulated collectionsof simple parts such as cylinders. Automatic initialization of such models has proved difficult and most approaches assume that the size and shape of the body parts are known a priori. In this paper we propose a method for automatically recovering a detailed parametric model of nonrigid body shape and pose from monocular imagery. Specifically, we represent the body using a parameterized triangulatedmesh model that is learned from a database of human range scans. We demonstrate a discriminative method to directly recover the model parameters frommonocular images using a conditional mixture of kernel regressors. This predicted pose and shape are used to initialize a generative model for more detailed pose and shape estimation. The resulting approach allows fully automatic pose and shape recovery from monocular and multi-camera imagery. Experimental resultsshow that our method is capable of robustly recovering articulated pose, shape and biometric measurements (e.g.


Key Cloud Computing Trends That Will Shape Enterprise Computing In 2020

#artificialintelligence

Cloud technology has many new things in store for the year 2020 and the future, a lot is happening and lot is also changing. In order to leverage the cloud's potential, organizations might need to maintain focus on leveraging AI advances and edge computing, this should be backed up with investing in sophisticated security to build and keep user trust. Cloud computing is on a great boom and there are innumerable trends to be discussed. It had amazing growth last year and is all set to break its own record in the next few. These days it's almost impossible to find an institution that doesn't rely to the least partially on cloud services.


Trix bringing back its popular fruity shapes: 'Kids of the 90s can rejoice'

FOX News

Trix is bringing back the beloved fruity shapes of the 90s. Silly Rabbit, Trix are for everyone who missed the iconic fruity shapes. General Mills Inc., the parent company of Trix, announced Monday it's bringing back the beloved cereal shape of the 90s and early 2000s after receiving thousands of requests from nostalgic fans. "In just the last 18 months, the brand has seen more than 20,000 requests, with fans asking things like'How many retweets to bring shapes back?' or'Is there a secret stash of Trix Shapes you can send me?'" Scott Baldwin, director of marketing for General Mills Cereal, said in a press release. "Kids of the 90s can rejoice, their fruity shapes are back in Trix."