no code implementations • 9 Feb 2024 • Ioannis N. Tzortzis, Konstantinos Makantasis, Ioannis Rallis, Nikolaos Bakalos, Anastasios Doulamis, Nikolaos Doulamis
We repeat the student training procedure by providing the assistance of the teacher model this time.
1 code implementation • 2 Feb 2024 • Chintan Trivedi, Nemanja Rašajski, Konstantinos Makantasis, Antonios Liapis, Georgios N. Yannakakis
In a more challenging setting, BehAVE manages to improve the zero-shot transferability of foundation models to unseen FPS games (up to 22%) even when trained on a game of a different genre (Minecraft).
no code implementations • 20 Jul 2023 • Chintan Trivedi, Konstantinos Makantasis, Antonios Liapis, Georgios N. Yannakakis
On-screen game footage contains rich contextual information that players process when playing and experiencing a game.
no code implementations • 18 May 2023 • Konstantinos Makantasis, Kosmas Pinitas, Antonios Liapis, Georgios N. Yannakakis
Privileged information enables affect models to be trained across multiple modalities available in a lab, and ignore, without significant performance drops, those modalities that are not available when they operate in the wild.
no code implementations • 14 Oct 2022 • Konstantinos Makantasis, Kosmas Pinitas, Antonios Liapis, Georgios N. Yannakakis
In particular, we assume that the ground truth of affect can be found in the causal relationships between elicitation, manifestation and annotation that remain \emph{invariant} across tasks and participants.
1 code implementation • 25 Aug 2022 • Kosmas Pinitas, Konstantinos Makantasis, Antonios Liapis, Georgios N. Yannakakis
Affect modeling is viewed, traditionally, as the process of mapping measurable affect manifestations from multiple modalities of user input to affect labels.
no code implementations • 5 Jul 2022 • Ioannis N. Tzortzis, Ioannis Rallis, Konstantinos Makantasis, Anastasios Doulamis, Nikolaos Doulamis, Athanasios Voulodimos
In Cultural Heritage, hyperspectral images are commonly used since they provide extended information regarding the optical properties of materials.
no code implementations • 4 Jul 2022 • Chintan Trivedi, Konstantinos Makantasis, Antonios Liapis, Georgios N. Yannakakis
Having access to accurate game state information is of utmost importance for any artificial intelligence task including game-playing, testing, player modeling, and procedural content generation.
1 code implementation • 20 Jun 2022 • Chintan Trivedi, Konstantinos Makantasis, Antonios Liapis, Georgios N. Yannakakis
Normalization is a vital process for any machine learning task as it controls the properties of data and affects model performance at large.
no code implementations • 13 Jun 2022 • Chintan Trivedi, Konstantinos Makantasis, Antonios Liapis, Georgios N. Yannakakis
We train an image encoder with three widely used SSL algorithms using solely the raw frames, and then attempt to recover the internal state variables from the learned representations.
no code implementations • 14 Apr 2022 • Kosmas Pinitas, Konstantinos Makantasis, Antonios Liapis, Georgios N. Yannakakis
Stochastic gradient descent (SGD) is a premium optimization method for training neural networks, especially for learning objectively defined labels such as image objects and events.
no code implementations • 12 Aug 2021 • Konstantinos Makantasis
Many of the affect modelling tasks present an asymmetric distribution of information between training and test time; additional information is given about the training data, which is not available at test time.
no code implementations • 11 Apr 2021 • Konstantinos Makantasis, Alexandros Georgogiannis, Athanasios Voulodimos, Ioannis Georgoulas, Anastasios Doulamis, Nikolaos Doulamis
We hereby propose the Rank-R Feedforward Neural Network (FNN), a tensor-based nonlinear learning model that imposes Canonical/Polyadic decomposition on its parameters, thereby offering two core advantages compared to typical machine learning methods.
no code implementations • 26 Jan 2021 • Konstantinos Makantasis, Antonios Liapis, Georgios N. Yannakakis
What if emotion could be captured in a general and subject-agnostic fashion?
Human-Computer Interaction
no code implementations • 17 Apr 2020 • Konstantinos Makantasis, Athanasios Voulodimos, Anastasios Doulamis, Nikolaos Bakalos, Nikolaos Doulamis
Recent advances in sensing technologies require the design and development of pattern recognition models capable of processing spatiotemporal data efficiently.
no code implementations • 10 Jul 2019 • Konstantinos Makantasis, Maria Kontorinaki, Ioannis Nikolos
To the best of our knowledge, this is one of the first approaches that propose a reinforcement learning driving policy for mixed driving environments.
no code implementations • 4 Jul 2019 • Konstantinos Makantasis, Antonios Liapis, Georgios N. Yannakakis
Is it possible to predict the affect of a user just by observing her behavioral interaction through a video?
no code implementations • 22 May 2019 • Konstantinos Makantasis, Maria Kontorinaki, Ioannis Nikolos
This work regards our preliminary investigation on the problem of path planning for autonomous vehicles that move on a freeway.
Robotics
no code implementations • 6 Feb 2019 • Konstantinos Makantasis, Anastasios Doulamis, Nikolaos Doulamis, Athanasios Voulodimos
In this work we propose a method for reducing the dimensionality of tensor objects in a binary classification framework.
no code implementations • 15 Feb 2018 • Konstantinos Makantasis, Anastasios Doulamis, Nikolaos Doulamis, Antonis Nikitakis, Athanasios Voulodimos
We also introduce a new learning algorithm to train the model, and we evaluate the \textit{Rank}-1 FNN on third-order hyperspectral data.
no code implementations • 24 Sep 2017 • Konstantinos Makantasis, Anastasios Doulamis, Nikolaos Doulamis, Antonis Nikitakis
Then, we introduce learning algorithms to train both the linear and the non-linear classifier in a way to i) to minimize the error over the training samples and ii) the weight coefficients satisfies the {\it rank}-1 canonical decomposition property.
no code implementations • 31 Jul 2016 • Konstantinos Makantasis, Antonis Nikitakis, Anastasios Doulamis, Nikolaos Doulamis, Yannis Papaefstathiou
Detection of moving objects in videos is a crucial step towards successful surveillance and monitoring applications.
1 code implementation • 2015 IEEE International Geoscience and Remote Sensing Symposium (IGARSS) 2015 • Konstantinos Makantasis, Konstantinos Karantzalos, Anastasios Doulamis, Nikolaos Doulamis
Our method exploits a Convolutional Neural Network to encode pixels' spectral and spatial information and a Multi-Layer Perceptron to conduct the classification task.
no code implementations • 29 Jun 2015 • Konstantinos Makantasis, Anastasios Doulamis, Nikolaos Doulamis
We propose a Gaussian mixture model for background subtraction in infrared imagery.