Istanbul Technical University

Signal Processing for Computational Intelligence Group


We develop algorithms to analyze contents of and compress signals from multitude sensors such as Visible/IR/Hyperspectral cameras, microphones, passive IR sensors, vibration sensor and spectrum sensors. We developed algorithms for hyperspectral data compression, forest fire, flame/smoke, volatile organic compound, falling person detection in video etc.
See book on Fire Detection


This project proposes a Markov model and wavelet transform-based technique to further improve the current state-of-the-art methods for video smoke detection by detecting signs of smoke existence in the MJPEG2000 compressed video. Existing video-based methods have been developed for the analysis of uncompressed Spatio-temporal sequences and are not appropriate for compressed video formats. Working with compressed video formats reduces the number of coefficients makes it suitable for parallel processing and paves the way for real-time processing which brings technical improvement in the field of video analysis. The proposed method is an appropriate alternative for traditional detection methods such as point and volumetric detectors and some of the computer vision-based approaches. For instance, point detectors do not help in large and spacious covered places; in the same manner of disadvantage, volumetric detectors issue an alarm when they «see» smoke within their viewing ranges. The proposed method has been applied on mostly home-made sequences as an initial result, and it will be carried out on the larger dataset in further works.

A presentation on smoke detection given at Purdue University.


We analyze tissue samples on a cellular basis where each cell is identified using membrane-based feature extraction and classification methods. Important research on this topic is myelin qualification in fluorescent microscopic images. Myelin sheath, wrapped around axons, allows rapid neural signal transmission, and degeneration of myelin causes various neurodegenerative diseases, such as Multiple Sclerosis (MS). For candidate drug discovery, it is essential to quantify myelin. This requires tedious expert labor comprising myelin labeling on microscopic fluorescence images, usually acquired by confocal microscopes. Detecting and quantifying myelin by using machine learning is the aim of this project.

Our poster on calcium signaling analysis for ALS diagnosis.


According to the report from GLOBOCAN 2018, 311,000 out of 570,000 cervical cancer cases have resulted in death. In the low human development index category, cervical cancer is the second most common cancer after breast cancer. In such an important type of cancer, early diagnosis is very important. Cervical squamous epithelial lesions (SIL) are cancer precursor lesions and their diagnosis is important because they have a chance of treatment before cancer development. Most of the grading studies in the literature are based on individual analysis of cell morphometry, which requires high computation time for optical zoom × 20 or × 40. In addition, pathologists can make different decisions for different cancer tissues and need objective confirmation. From this point of view, we present two statistical properties extraction methods for the classification of hematoxylin and eosin-stained cervical tissues: Local Histogram Features (LHF) and Nuclei Density Features (NDF). Our method is based on dividing the tissues into basal, central, and upper segments and using the histograms of each segment as a feature. As a similarity measurement, the symmetric Kullback-Leibler metric is planned to be used among LHF feature vectors.


We develop algorithms to compress hyperspectral images using tools such as discrete wavelet transforms and graph signal processing. Hyperspectral images are images volumetric images with composed of a continuous bands with diferent wavelengths. In our research, we have developed ways to comress the images using different techniques such as appying different vavelets with lookup tables and graph singal processing.

Our poster on hyperspectral image compression.

Our poster on smoke detection from H264 compressed videos.


Artificial Neural Networks are addressing major problems of computer vision, natural language processing, and data science in the last decade with increasing computational power and amount of data. Despite its popularity and success, when presented with a sequence of tasks with only having access to the current task's data, neural networks fail to preserve their performance on previously learned tasks. This problem is called catastrophic forgetting and one of the biggest obstacles on the way of artificial general intelligence. The purpose of continual learning is to be able to learn a sequence of tasks without suffering from catastrophic forgetting while improving forward and backward transfer across tasks.

Obtaining a high resolution (HR) image from the low resolution (LR) image is the goal of single image super-resolution (SISR). One of the crucial points is the difficulty of finding missing high-frequency details. The second is how LR images and HR images correspond to each other. Developing a patch-based deep learning super-resolution method is the aim of this project.

Deep learning methods outperform traditional methods on problems of the computer vision field. However, these methods require high computational cost, process time and memory, since they consist of large numbers of parameters. Using these methods is impossible on devices with low-budget and real-time applications. Hence, model compression has become an emerging topic, which aims to decrease the costs while achieving the least loss of performance. In this scope, we are trying to develop novel model compression methods and examine them both theoretically and practically.

In our study, we focus on knowledge distillation which is one of the model compression techniques. It basically means transferring knowledge from teacher models to student models. In our work, PURSUhInT: In Search of Informative Hint Points Based on Layer Clustering for Knowledge Distillation, we propose a method for determining efficient hint positions that teacher's knowledge is taken from. For this purpose, we obtain clusters of layers from pretrained teacher models and select one representative layer for each cluster as hint positions in order to avoid redundancy in terms of layers' knowledge.