Rozdziały

Stały URI dla kolekcjihttp://hdl.handle.net/11652/4775

Przeglądaj

Filtry

20234

Ustawienia

Autorsearch.filters.applied.operator.equals: search.filters.author.Ślot, KrzysztofMa przypisane plikisearch.filters.applied.operator.equals: Tak

collection.search.results.head

Teraz wyświetlane 1 - 4 z 4
  • Pozycja
    Grounded HyperSymbolic Representations Learned through Gradient-Based Optimization
    (Wydawnictwo Politechniki Łódzkiej, 2023) Łuczak, Piotr; Ślot, Krzysztof; Kucharski, Jacek
    Hyperdimensional computing is a novel paradigm, capable of processing complex data structures with simple operations. Its main limitations lie in the conversion of real world data onto hyperdimensional space, which due to lack of a universal translation scheme, oftentimes requires application-specific methods. This work presents a novel method for unsupervised hyperdimensional conversion of arbitrary image data. Additionally, this method is augmented by the ability of creating HyperSymbols, or class prototypes, provided that such class labels are available. The proposed method achieves promising performance on MNIST dataset, both in translating individual samples as well as producing HyperSymbols for downstream classification task.
  • Pozycja
    Autoregressive Label-Conditioned Autoencoder for Controllable Image-To-Video Generation
    (Wydawnictwo Politechniki Łódzkiej, 2023) Kubicki, Kacper; Ślot, Krzysztof
    Generating videos from a single image with user-controlled attributes is a complex challenge in the field of computer vision, despite the significant advancements recently made in the field. This paper presents a novel approach to tackle this issue, leveraging a convolutional autoencoder with supervised principal component analysis and autoregressive inference step. The efficacy of the proposed method is evaluated on two datasets – MNIST handwritten-digits and time-lapse photos of the sky. Results from both quantitative and qualitative analyses show that the proposed approach produces high-quality videos of variable duration with user-defined attributes, while preserving the integrity of original image contents.
  • Pozycja
    RNN-based Phase Unwrapping for Enabling Vital Parameter Monitoring with FMCW Radars
    (Wydawnictwo Politechniki Łódzkiej, 2023) Łuczak, Piotr; Hausman, Sławomir; Ślot, Krzysztof
    Application of radar technology enables remote breathing and heart rate monitoring by analyzing motion waveforms, which are reconstructed from phase signals extracted from radar-delivered data. However, nonlinear deformations introduced by phase recovery procedure make accurate motion reconstruction highly challenging, especially for millimeter-long waves that are commonly generated by state-of-the-art radar devices. In the presented paper we show that a GRU-based neural predictor is capable of correct phase unwrapping under presence of noise (originating e.g. from random subject’s movements), enabling vital parameter monitoring in realistic scenarios, which cannot be accomplished using standard approaches.
  • Pozycja
    Towards Detection of Unknown Polymorphic Patterns Using Prior Knowledge
    (Wydawnictwo Politechniki Łódzkiej, 2023) Kucharski, Przemysław; Ślot, Krzysztof
    The presented paper proposes a novel approach for detecting unknown polymorphic patterns in sequences composed of random symbols and of known polymorphic patterns. We propose to represent rules that drive pattern generation as regular expressions. To detect unknown patterns, we first incorporate knowledge on known rules into a Convolutional Autoencoder (CAE), then we train the CAE with additional objective to prevent weights from learning the already known patterns. Analysis of training results provides statistically significant information on presence or absence of polymorphic patterns that were not previously known.