Or Dinari

I am currently researching Computer Vision at Apple. I completed my PhD in Computer Science under the supervision of Dr. Oren Freifeld, focusing on unsupervised/semi-supervised learning, clustering, Bayesian nonparametrics, deep learning, and computer-vision applications.

Publications

1. Revisiting DP-Means: Fast Scalable Algorithms via Parallelism and Delayed Cluster Creation. Or Dinari and Oren Freifeld, UAI 2022.
   [Paper] [Code]


2. Variational- and Metric-based Deep Latent Space for Out-of-Distribution Detection. Or Dinari and Oren Freifeld, UAI 2022.
   [Paper] [Code]


3. Unsupervised Online Multiple Animal Tracking. Dor Litvak, Or Dinari and Oren Freifeld, CV4Animals Workshop, CVPR 2022.
   [Paper - Coming Soon] [Code -Coming Soon]


4. Sampling in Dirichlet Process Mixture Models for Clustering Streaming Data. Or Dinari and Oren Freifeld, AISTATS 2022.
   [Paper] [Code - Julia] [Code - Python] [DataSets] [Example Notebook]


5. Common Failure Modes of Subcluster-based Sampling in Dirichlet Process Gaussian Mixture Models - and a Deep-learning Solution. Vlad Winter*, Or Dinari* and Oren Freifeld (* Indicates that both these authors contibuted equally), AISTATS 2022.
   [Paper] [Code]


6. CPU-and GPU-based Distributed Sampling in Dirichlet Process Mixtures for Large-scale Analysis Or Dinari*, Raz Zamir*, J. W. Fisher III and Oren Freifeld (* Indicates that both these authors contibuted equally), Under Review.
   [Paper] [Code]


7. Scalable and Flexible Clustering of Grouped Data via Parallel and Distributed Sampling in Versatile Hierarchical Dirichlet Processes. Or Dinari and Oren Freifeld, UAI 2020.
   [Paper] [Supplemental] [Code - Julia] [Code - Python]


8. Distributed MCMC inference in Dirichlet process mixture models using Julia. Or Dinari*, Angel Yu*, Oren Freifeld, and J. W. Fisher III (* Indicates that both these authors contibuted equally). HPML 2019.
   [Paper] [Slides] [Preprint] [Code] [Python wrapper]
   Best Paper Award
   

Code

DPMMSubClustersStreaming.jl

Julia package for clustering streaming data using DPMM.
Easy to use and configure.
Code for our paper Sampling in Dirichlet Process Mixture Models for Clustering Streaming Data,

dpmmpythonStreaming

Python wrapper for the DPMMSubClusterStreaming.jl package.

DPMMSubClusters.jl

Julia package for distributed parallel inference in Dirichlet Process Mixture Models. Easy to use and configure.
Code for our paper Distributed MCMC Inference in Dirichlet Process Mixture Models Using Julia

dpmmpython

Python wrapper for the DPMMSubClusters.jl package.

VersatileHDPMixtureModels.jl

This package allows to perform inference in the vHDPMM setting, as described in our paper Scalable and Flexible Clustering of Grouped Data via Parallel and Distributed Sampling in Versatile Hierarchical Dirichlet Processes, or as an alternative, it can perform inference in HDPMM setting.

Explore ImageNet

Small python package to explore the hierarchical struct of ImageNet/WordNet.

Teachings

Contact Me