Our paper, “Semantic Segmentation of Crop Areas in Remote Sensing Imagery using Spectral Indices and Multiple Channels” has been accepted for publication in the ICMV2022!

Abstract

This study focuses on pixel-wise semantic segmentation of crop production regions by using satellite remote sensing multispectral imagery. One of the principal aims of the study is to find out whether the raw multiple channel inputs are more effective in the training process of the semantic segmentation models or if the formularised counterparts as the spectral indices are more effective. For this purpose, the vegetation indices NDVI, ARVI and SAVI and the water indices NDWI, NDMI, and WRI are employed as inputs. Additionally, using 8, 10 and 16 channels, multiple channel inputs are utilised. Moreover, all spectral indices are taken as separate channels to form a multiple channel input.  We conduct deep learning experiments using two semantic segmentation architectures, namely U-Net and DeepLabV3+. Our results show that, in general, feeding raw multiple channel inputs to semantic segmentation models performs much better than feeding the spectral indices. Hence, regarding crop production region segmentation, deep learning models are capable of encoding multispectral information. The results also reveal that spatial resolution of multispectral data has a significant effect on the semantic segmentation performance, and therefore the RGB band, which has the lowest ground sample distance (0.31 m) outperforms multispectral bands and shortwave infrared bands.

The official link can be found here: https://www.spiedigitallibrary…

Our paper entitled “Augmentation of Atmospheric Turbulence Effects on Thermal Adapted Object Detection Models” has been accepted for publication in the CVPR2022 Perception Beyond the Visible Spectrum (PBVS) Workshop. Anybody interested in the title should definitely take a look!

Abstract

Atmospheric turbulence has a degrading effect on the image quality of long-range observation systems. As a result of various elements such as temperature, wind velocity, humidity, etc., turbulence is characterized by random fluctuations in the refractive index of the atmosphere. It is a phenomenon that may occur in various imaging spectra such as the visible or the infrared bands. In this paper, we analyze the effects of atmospheric turbulence on object detection performance in thermal imagery. We use a geometric turbulence model to simulate turbulence effects on a medium-scale thermal image set, namely “FLIR ADAS v2”. We apply thermal domain adaptation to state-of-the-art object detectors and propose a data augmentation strategy to increase the performance of object detectors which utilizes turbulent images in different severity levels as training data. Our results show that the proposed data augmentation strategy yields an increase in performance for both turbulent and non-turbulent thermal test images.

The official link can be found here: ieeexplore.ieee.org…,
and the arXiv preprint is here: https://arxiv.org/abs/2204.08745

Our paper entitled “A Survey on Deep Learning-based Architectures for Semantic Segmentation on 2D Images” has been accepted for publication in the Journal of Applied Artificial Intelligence. This was a collaborative study with Dr. Ulku from Ankara University. We did our best to summarize the journey of deep semantic segmentation. Anybody interested in the title should definitely take a look!

Abstract

Semantic segmentation is the pixel-wise labeling of an image. Boosted by the extraordinary ability of convolutional neural networks (CNN) in creating semantic, high-level and hierarchical image features; several deep learning-based 2D semantic segmentation approaches have been proposed within the last decade. In this survey, we mainly focus on the recent scientific developments in semantic segmentation, specifically on deep learning-based methods using 2D images. We started with an analysis of the public image sets and leaderboards for 2D semantic segmentation, with an overview of the techniques employed in performance evaluation. In examining the evolution of the field, we chronologically categorized the approaches into three main periods, namely pre-and early deep learning era, the fully convolutional era, and the post-FCN era. We technically analyzed the solutions put forward in terms of solving the fundamental problems of the field, such as fine-grained localization and scale invariance. Before drawing our conclusions, we present a table of methods from all mentioned eras, with a summary of each approach that explains their contribution to the field. We conclude the survey by discussing the current challenges of the field and to what extent they have been solved.

The official link can be found here: https://www.tandfonline…,
and the arXiv preprint is here: http://arxiv.org/abs/1912.10230

Starting with 2021-2022 Fall Semester
I will be lecturing a new course:
“MMI711 – Sequence Models in Multimedia”

MMI711 will cover various concepts related to understanding and processing different types of
multimedia sequence models. The course starts with an overview of sequence models, recurrent neural networks (RNN) and continues with details on training RNNs. By introducing different sequence modelling problems, recurrent architectural models and variants of gated units the course covers all fundamental concepts related to sequence learning in intelligent multimedia systems. In addition the course covers the recurrent and nonrecurrent models of attention in various multimedia type signals such as vision and/or sound.

During the registrations, MMI711 will only be available for MMI program students of the Informatics Institute. At this point, we do not know if there is going to be a sufficient quota for other departments’ students. If you are interested and would like to be lined up for a possible quota, please fill this form (registrations closed).

Only during the add-drops, we will be able to tell you if you will be officially registered for this course. The first two weeks of the course will be online in Zoom, accepting any visitors.

Starting with 2021-2022 Fall Semester,
I will be lecturing a new course:
“DI504 – Foundations of Deep Learning”

DI504 will cover various subjects related to fundamental concepts in deep learning. The course starts with the introductory concepts in deep learning, such as neural networks, training, high level features and performance evaluations. In the following, the course covers different application fields of deep learning, to what extent the methods succeed and what the next step is.

During the registrations, DI504 will only be available for DI program and MMI program students of the Informatics Institute. At this point, we do not know if there is going to be a sufficient quota for other departments’ students. If you are interested and would like to be lined up for a possible quota, please fill this form (registrations closed).

Only during the add-drops, we will be able to tell you if you will be officially registered to this course. The first two weeks of the course will be online in Zoom, accepting any visitors.

Our paper entitled “Dynamical System Parameter Identification using Deep Recurrent Cell Networks” has been accepted for publication in Neural Computing and Applications. This was a collaborative study with Dr. Cifdaloz from Çankaya University EEE. This is a (hopefully many of the first) paper on deep learning-based control engineering.

Abstract

In this paper, we investigate the parameter identification problem in dynamical systems through a deep learning approach. Focusing mainly on second-order, linear time-invariant dynamical systems, the topic of damping factor identification is studied. By utilizing a six-layer deep neural network with different recurrent cells, namely GRUs, LSTMs or BiLSTMs; and by feeding input/output sequence pairs captured from a dynamical system simulator, we search for an effective deep recurrent architecture in order to resolve the damping factor identification problem. Our study’s results show that, although previously not utilized for this task in the literature, bidirectional gated recurrent cells (BiLSTMs) provide better parameter identification results when compared to unidirectional gated recurrent memory cells such as GRUs and LSTM. Thus, indicating that an input/output sequence pair of finite length, collected from a dynamical system and when observed anachronistically, may carry information in both time directions to predict a dynamical systems parameter.

The official link can be found here: https://rdcu.be/cn8jY,
and the arXiv preprint is here: http://arxiv.org/abs/2107.02427