NewEvery arXiv paper, its researchers & institutions — mapped.
the archive

#probabilistic modeling

11 results
cs.CV2019

Visual Dynamics: Stochastic Future Generation via Layered Cross Convolutional Networks

Tianfan Xue, Jiajun Wu, Katherine L. Bouman +1

The paper introduces a probabilistic model using a Cross Convolutional Network to generate multiple plausible future video frames from a single input image.

#future frame prediction#probabilistic modeling#cross convolutional networks#video synthesis
cs.LG2019

Random Sum-Product Forests with Residual Links

Fabrizio Ventola, Karl Stelzner, Alejandro Molina +1

The paper proposes Random Sum-Product Forests (RSPFs), an ensemble of randomly generated sum‑product networks, and introduces residual links that let components share specialized s…

#sum-product networks#ensemble methods#residual connections#density estimation
cs.CV2019

Probabilistic Face Embeddings

Yichun Shi, Anil K. Jain

The paper introduces Probabilistic Face Embeddings, which model each face image as a Gaussian distribution in a latent space to capture both feature estimates and their uncertainty…

#face recognition#embedding#probabilistic modeling#uncertainty estimation
eess.IV2019

Confident Head Circumference Measurement from Ultrasound with Real-time Feedback for Sonographers

Samuel Budd, Matthew Sinclair, Bishesh Khanal +6

The paper presents a probabilistic deep learning system that provides real-time feedback on the confidence of fetal head circumference measurements from ultrasound, helping sonogra…

#fetal ultrasound#head circumference measurement#real-time feedback#deep learning
eess.AS2019

Maximum likelihood convolutional beamformer for simultaneous denoising and dereverberation

Tomohiro Nakatani, Keisuke Kinoshita

The paper proposes a probabilistic formulation of the Weighted Power minimization Distortionless response (WPD) convolutional beamformer, unifying weighted prediction error derever…

#beamforming#dereverberation#denoising#maximum likelihood
eess.AS2019

Probabilistic Permutation Invariant Training for Speech Separation

Midia Yousefi, Soheil Khorram, John H. L. Hansen

The paper introduces Probabilistic Permutation Invariant Training (Prob‑PIT), a method that treats the output‑label assignment as a latent random variable to improve speaker‑indepe…

#speech separation#permutation invariant training#probabilistic modeling#deep learning
cs.AI2019

Constructing High Precision Knowledge Bases with Subjective and Factual Attributes

Ari Kobren, Pablo Barrio, Oksana Yakhnenko +2

The paper proposes a probabilistic consensus modeling approach, using neural networks, to build knowledge bases that store both factual and subjective attributes while allowing the…

#knowledge base construction#subjective attribute modeling#precision control#probabilistic modeling
cs.CV2019

Conditional Generative Neural System for Probabilistic Trajectory Prediction

Jiachen Li, Hengbo Ma, Masayoshi Tomizuka

The paper introduces a conditional generative neural system that predicts multiple plausible future trajectories for dynamic agents by modeling the distribution of possible motions…

#trajectory prediction#probabilistic modeling#conditional generative models#autonomous vehicles
cs.LG2019

Probabilistic Models of Relational Implication

Xavier Holt

The paper introduces a principled probabilistic framework for learning when one relation implies another in knowledge bases, achieving higher AUC scores than prior methods and prov…

#relational implication#knowledge bases#probabilistic modeling#link prediction
eess.IV2019

PHiSeg: Capturing Uncertainty in Medical Image Segmentation

Christian F. Baumgartner, Kerem C. Tezcan, Krishna Chaitanya +6

The paper introduces a hierarchical probabilistic model that captures uncertainty in medical image segmentation by using latent variables at multiple resolutions, trained via a var…

#image segmentation#uncertainty quantification#probabilistic modeling#variational autoencoders
stat.AP2019

Decision Tree Learning for Uncertain Clinical Measurements

Cecília Nunes, Hélène Langet, Mathieu De Craene +3

The paper introduces a probabilistic decision‑tree method that explicitly models measurement noise during split‑threshold selection, training instance splitting, and prediction, sh…

#decision trees#uncertain data#probabilistic modeling#clinical measurements