Leo F. Isikdogan, PhD
Researcher • Engineer • Artist
Deep Learning Algorithm Engineer at Apple Inc.

I work as a Deep Learning Algorithm Engineer at Apple. Previously, I worked for Intel Corporation and Motorola/Lenovo.

I received my PhD from The University of Texas at Austin, where I worked with Alan Bovik and Paola Passalacqua.

I make educational videos on machine learning, computer vision, and image processing on my YouTube channel.

I use artificial intelligence and creative algorithms to experiment with different styles of art. Check out my artwork on Instagram!

Projects & Publications

Classical Music + AI Art

2020 • Personal Project | A generative music visualizer. Supports up to 8K resolution.

YouTube Channel
MARCUS: An AI Art Generator

2020 • Personal Project | An AI-powered software tool for experimenting with different styles of art.

SemifreddoNets: Partially Frozen Neural Networks for Efficient Computer Vision Systems

ECCV 2020 | A system comprised of fixed-topology neural networks having partially frozen weights.

Arxiv Springer PDF Video
A Machine Learning Imaging Core using Separable FIR-IIR Filters

2020 | Fixed-function neural network hardware that performs pixel-to-pixel image transformations in a highly efficient way.

Arxiv PDF Video
Eye Contact Correction using Deep Neural Networks

WACV 2020 | Restores eye contact during video calls by warping and tuning the eyes in the input frames.

Arxiv PDF (CVF Open Access) Video | Press coverage: Fortune TechXplore Edgy.app VC Daily test.ai
Characterization of Deltaic Channel Morphodynamics From Imagery Time Series Using the Channelized Response Variance

JGR 2019 (Journal) | Follow-up study on RivaMap. Proposes a new metric for tracking temporal change in channel systems.

Wiley Online Library PDF
Seeing Through the Clouds With DeepWaterMap

GRSL 2019 (Letter) | The state-of-the-art convolutional neural network for pixel-wise water segmentation.

IEEEXplore PDF Code
VisionISP: an Image Processing Pipeline for Computer Vision Applications

ICIP 2019 | An imaging pipeline that consists of trainable blocks.

IEEEXplore Arxiv PDF Video
Learning a River Network Extractor Using an Adaptive Loss Function

GRSL 2018 (Letter) | A deep convolutional neural network that detects river centerlines on satellite imagery.

IEEEXplore PDF Code
Landmark Recognition on Moto X4 Smart Camera

2017 | Detects points of interest in a picture and asks if you want to learn more about them.

Video | Press coverage: Qualcomm Blog Motorola Blog Android Headlines
Surface Water Mapping by Deep Learning

IEEE JSTARS 2017 (Journal) | A deep model that segments water on multispectral images.

IEEEXplore PDF Code
RivaMap: An Automated River Analysis and Mapping Engine

Elsevier RSE 2017 (Journal) | Processes Landsat images at large scales, delineates rivers, and estimates their width.

ScienceDirect PDF Code
A New Database and Protocol for Image Reuse Detection

ECCV Workshop on Computer Vision for Art Analysis 2016 | Image reuse detection in digital artworks.

Springer PDF
Automatic Channel Network Extraction From Remotely Sensed Images by Singularity Analysis

GRSL 2015 (Letter) | Hand-crafted features to detect curvilinear structures on remotely sensed images.

IEEEXplore PDF Code
UTimelapse: a Tool for Creating High-Quality Timelapse Videos

2014 | Creates high-quality timelapse videos by reducing camera shake, flickering, and single-frame artifacts.

Blog Post Video Google Scholar
Affine Invariant Salient Patch Descriptors for Image Retrieval

WIAMIS 2013 | A foreground-sensitive image descriptor.

A Real Time Virtual Dressing Room Application

2012 | Used a Microsoft Kinect sensor to superimpose t-shirts on a user.

Technical Report Video
Automatic Recognition of Turkish Fingerspelling

INISTA 2011 | A sign language recognition app that translated fingerspelled letters into text.