Hi, I’m Tal!

I’m a Ph.D. candidate at the Technion, Israel Institute of Technology, in the Electrical and Computer Engineering (ECE) Department, under the supervision of Prof. Aviv Tamar. I’m a member of the RL^2 lab and the Control, Robotics and Machine Learning (CRML) lab.

I received my B.Sc. (Cum Laude) and M.Sc. (Summa Cum Laude) both in Electrical and Computer Engineering from the Technion.

My research interests include (Deep) Unsupervised/Self-supervised Representation Learning, Generative Modeling, Reinforcment Learning and Robotics.

During my time at the Technion, I’m also a teaching assistant in Machine Learning courses, for which I have written theory-based and hands-on tutorials that can be accessed under the “Teaching” section.

I’m grateful and honored to be supported by
The Irwin and Joan Jacobs 2024 Ph.D. Excellence Fellowship
The Miriam and Aaron Gutwirth Memorial 2023 Ph.D. Excellence Fellowship

You can contact me at taldanielm at campus dot technion dot ac dot il.



Publications and Pre-Prints

DDLP: Unsupervised Object-centric Video Prediction with Deep Dynamic Latent Particles
Tal Daniel and Aviv Tamar
Transactions on Machine Learning Research (TMLR) 2024
TL;DR - A new object-centric video prediction, generation and modification algorithm based on the deep latent particle (DLP) representation.

Entity-Centric Reinforcement Learning for Object Manipulation from Pixels
Dan Haramati, Tal Daniel and Aviv Tamar
International Conference on Learning Representations (ICLR) 2024 Spotlight
TL;DR - We use deep latent particles (DLP) as input instead of pixels and unlock compositional capabilities in reinforcement learning.

Unsupervised Image Representation Learning with Deep Latent Particles
Tal Daniel and Aviv Tamar
International Conference on Machine Learning (ICML) 2022
TL;DR - Represent images as latent particles, unsupervised detection and segmentation of objects, particles can be controlled and used for image manipulation and video prediction.

Soft-IntroVAE: Analyzing and Improving the Introspective Variational Autoencoder
Tal Daniel and Aviv Tamar
IEEE / CVF Computer Vision and Pattern Recognition Conference (CVPR) 2021 Oral
MLIS-TCE Conference 2022 - Best Paper Award
TL;DR - Stable adversarial training of VAEs without a discriminator, applicable for density estimation, image generation, image translation, Out-of-Distribution detection and many more.

Deep Variational Semi-Supervised Novelty Detection
Tal Daniel, Thanard Kurutach and Aviv Tamar
NeurIPS 2021 Workshop on Deep Generative Models and Downstream Applications
TL;DR - Principled incorporation of negative samples in the VAE framework for meaningful representations.

Beyond Credential Stuffing: Password Similarity Models Using Neural Networks
Bijeeta Pal, Tal Daniel, Rahul Chatterjee and Thomas Ristenpart
2019 IEEE Symposium on Security and Privacy (SP)
TL;DR - Cracking passwords with neural networks, but also defend against such attacks.



I teach several courses at the Technion, and the materials are available on GitHub to everyone. All the tutorials are in a Jupyter Notebook format and include theory and code in Python and PyTorch. PDF version is also available.

ECE - 046211 - Deep Learning
Topics: Single Neuron, PyTorch Basics, Optimization and Gradient Descent-based Algorithms, Automatic Differentiation (AutoDiff) and PyTorch's AutoGrad, Multilayer Neural Networks, Convolutional Neural Networks (CNNs), Sequential Tasks, Recureent Neural Networks (RNNs), Attention, Transformer, Training Methods, Bayesian Hyper-parameter Tuning with Optuna, Transfer Learning, Reperesentation and Self-Supervised Learning
[GitHub] [Students' Projects Website] [Video Tutorials]

ECE - 046746 - Computer Vision
Topics: Image Processing Basics, PyTorch Basics, 2D Convolution, Convolutional Neural Networks, (Deep) Semantic Segmentation, (Deep) Object Detection, (Deep) Object Tracking, Generative Adversarial Network (GAN), 3D Deep Learning Basics
Spring 2020 (with Dahila Urbach), Spring 2021 (with Elias Nehme)

ECE - 046202 - Unsupervised Learning and Data Analysis
Topics: Statistics (estimators, confidence intervals, hypothesis testing), Dimensionality Reduction (PCA, KPCA, t-SNE), Deep Generative Models (VAE, GAN), Clustering (K-Means, EM algorithm, Spectral Clustering)
Winter 2020, Winter 2021

CS - 236756 - Introduction to Machine Learning
Topics: Probability and Linear Algebra Basics, PCA, Feature Selection, Evaluation and Validation methods, Optimization, Decision Trees, Linear Regression, Linear Classifiers, EM algorithm, Boosting and Bagging, SVM, Deep Learning introduction, PAC Learning
Spring 2019, Spring 2020


Other Projects and Cool Stuff

Interactive GUI for Deep (Dynamic) Latent Particels (D(D)LP)

[Project Site]

Interpolation between airplane and car in the latent space of 3D Soft-IntroVAE

[Project Site]

Python Implementation of Pencil Drawing by Sketch and Tone (Lu et al., NPAR 2012)


PyTorch implementation of Least-Squares DQN (Levine, Zahavy et al, NeurIPS 2017) with extras (DuelingDQN, Boosted FQI)


Bayesian Gradient Descent Algorithm (Zeno et al, 2019) Model for TensorFlow


Android application for mining Scrypt coins (Litecoin) with custom options.

With Eyal Ben-David
[GitHub][Simple Stratum Miner]



You can contact me at taldanielm at campus dot technion dot ac dot il.