We present a latent diffusion model over 3D scenes, that can be trained using only 2D
image data. To achieve this, we first design an autoencoder that maps multi-view images to …

The standard approach to tackling computer vision problems is to train deep convolutional
neural network (CNN) models using large-scale image datasets which are representative of …

The rapid evolution of wireless technologies and applications has catapulted us into an era
where real-time data processing and low latency are more critical than ever. With significant …

We present a method for generating alpha mattes using a limited data source. We pretrain a
novel transformerbased model (StyleMatte) on portrait datasets. We utilize this model to …

Neural 3D scene representations have recently emerged as a novel way to store information
about 3D environments, their properties, and behavior, with learned features. Their key …

Deep learning, particularly generative models, has inspired controllable image synthesis
methods and applications. These approaches aim to generate specific visual content using …

High fidelity reconstruction and editing of objects is a challenging task in the graphics and
vision community. Recent work in 3D reconstruction are unable to preserve high-frequency …

Desde la aparición del Deep Learning, cada vez se necesitan más datos para el
entrenamiento de modelos de inteligencia artificial. Este problema se está solucionando …

Our objective is to enhance the generalization capabilities of existing machine perception
models and achieve diverse domain alignments through adept representation learning …

With the vast spread of automation, the significant demand for communication services, the
addition of IoT to the Internet, and the corresponding increase and sophistication in …