-
PTSD-MDNN : Fusion tardive de réseaux de neurones profonds multimodaux pour la détection du trouble de stress post-traumatique
Authors:
Long Nguyen-Phuoc,
Renald Gaboriau,
Dimitri Delacroix,
Laurent Navarro
Abstract:
In order to provide a more objective and quicker way to diagnose post-traumatic stress disorder (PTSD), we present PTSD-MDNN which merges two unimodal convolutional neural networks and which gives low detection error rate. By taking only videos and audios as inputs, the model could be used in the configuration of teleconsultation sessions, in the optimization of patient journeys or for human-robot…
▽ More
In order to provide a more objective and quicker way to diagnose post-traumatic stress disorder (PTSD), we present PTSD-MDNN which merges two unimodal convolutional neural networks and which gives low detection error rate. By taking only videos and audios as inputs, the model could be used in the configuration of teleconsultation sessions, in the optimization of patient journeys or for human-robot interaction.
△ Less
Submitted 14 March, 2024;
originally announced March 2024.
-
M&M: Multimodal-Multitask Model Integrating Audiovisual Cues in Cognitive Load Assessment
Authors:
Long Nguyen-Phuoc,
Renald Gaboriau,
Dimitri Delacroix,
Laurent Navarro
Abstract:
This paper introduces the M&M model, a novel multimodal-multitask learning framework, applied to the AVCAffe dataset for cognitive load assessment (CLA). M&M uniquely integrates audiovisual cues through a dual-pathway architecture, featuring specialized streams for audio and video inputs. A key innovation lies in its cross-modality multihead attention mechanism, fusing the different modalities for…
▽ More
This paper introduces the M&M model, a novel multimodal-multitask learning framework, applied to the AVCAffe dataset for cognitive load assessment (CLA). M&M uniquely integrates audiovisual cues through a dual-pathway architecture, featuring specialized streams for audio and video inputs. A key innovation lies in its cross-modality multihead attention mechanism, fusing the different modalities for synchronized multitasking. Another notable feature is the model's three specialized branches, each tailored to a specific cognitive load label, enabling nuanced, task-specific analysis. While it shows modest performance compared to the AVCAffe's single-task baseline, M\&M demonstrates a promising framework for integrated multimodal processing. This work paves the way for future enhancements in multimodal-multitask learning systems, emphasizing the fusion of diverse data types for complex task handling.
△ Less
Submitted 14 March, 2024;
originally announced March 2024.
-
Toward More Accurate and Robust Optimal Power Flow for Distribution Systems
Authors:
Dakota Hamilton,
Loraine Navarro,
Dionysios Aliprantis
Abstract:
The objective of this paper is to improve the accuracy and robustness of optimal power flow (OPF) formulations for distribution systems modeled down to the low-voltage point of connection of individual buildings. An approach for addressing the uncertain switching behavior of building loads(e.g., air conditioners, water heaters, or pool pumps) and variable renewable generation (e.g., rooftop solar)…
▽ More
The objective of this paper is to improve the accuracy and robustness of optimal power flow (OPF) formulations for distribution systems modeled down to the low-voltage point of connection of individual buildings. An approach for addressing the uncertain switching behavior of building loads(e.g., air conditioners, water heaters, or pool pumps) and variable renewable generation (e.g., rooftop solar) in the OPF is proposed. Rather than using time-averaged forecasts to determine voltage magnitude constraints, we leverage worst-case minimum and maximum forecasts of loads and distributed energy resource generation. Sensitivities of the power flow equations are used to predict how these deviations in load and renewable generation will impact system voltages, and the voltage constraints in the OPF are dynamically adjusted to mitigate voltage violations due to this uncertainty. A methodology for incorporating models of split-phase components and transformer core losses in the OPF formulation is also proposed. The proposed approach is validated through numerical case studies on a realistic distribution feeder using GridLAB-D, a distribution system simulation software.
△ Less
Submitted 9 October, 2023;
originally announced October 2023.
