With the number of people with dementia (PwD) increasing worldwide, the need for support for caregivers is growing. However, the burden on caregivers is exceptionally high for care related to elimination. Because toileting is generally performed in private rooms, caregivers cannot share their knowledge with other caregivers. Therefore, this study focused on creating a care suggestion system that generates care methods based on expert knowledge by extracting highly relevant information using retrieval-augmented generation (RAG) and providing it to a generative pretrained transformer (GPT) using knowledge regarding toileting care for people with dementia (PwD). The documents containing knowledge of care for PwD and the questions were vectorized, and prompts were provided with documents that were similar to the questions. Three types were prepared: a case in which only GPT was used, a case in which all information was provided, and a case in which the proposed method was used. To evaluate the proposed method, an experiment was conducted with nurses with experience in dementia nursing or gerontological nursing care. Five items were set for the three types of responses generated, and each item was rated on a 5-point scale. The results indicated the usefulness of using RAGs to obtain suggestions for care methods.

Demographic bias is one of the major challenges for face recognition systems. The majority of existing studies on demographic biases are heavily dependent on specific demographic groups or demographic classifier, making it difficult to address performance for unrecognised groups. This paper introduces "LabellessFace", a novel framework that improves demographic bias in face recognition without requiring demographic group labeling typically required for fairness considerations. We propose a novel fairness enhancement metric called the class favoritism level, which assesses the extent of favoritism towards specific classes across the dataset. Leveraging this metric, we introduce the fair class margin penalty, an extension of existing margin-based metric learning. This method dynamically adjusts learning parameters based on class favoritism levels, promoting fairness across all attributes. By treating each class as an individual in facial recognition systems, we facilitate learning that minimizes biases in authentication accuracy among individuals. Comprehensive experiments have demonstrated that our proposed method is effective for enhancing fairness while maintaining authentication accuracy.

The paper provides a summary of the 2023 Unconstrained Ear Recognition Challenge (UERC), a benchmarking effort focused on ear recognition from images acquired in uncontrolled environments. The objective of the challenge was to evaluate the effectiveness of current ear recognition techniques on a challenging ear dataset while analyzing the techniques from two distinct aspects, i.e., verification performance and bias with respect to specific demographic factors, i.e., gender and ethnicity. Seven research groups participated in the challenge and submitted a seven distinct recognition approaches that ranged from descriptor-based methods and deep-learning models to ensemble techniques that relied on multiple data representations to maximize performance and minimize bias. A comprehensive investigation into the performance of the submitted models is presented, as well as an in-depth analysis of bias and associated performance differentials due to differences in gender and ethnicity. The results of the challenge suggest that a wide variety of models (e.g., transformers, convolutional neural networks, ensemble models) is capable of achieving competitive recognition results, but also that all of the models still exhibit considerable performance differentials with respect to both gender and ethnicity. To promote further development of unbiased and effective ear recognition models, the starter kit of UERC 2023 together with the baseline model, and training and test data is made available from: http://ears.fri.uni-lj.si/

In this paper, we propose an office layout plan evaluation system using evacuation simulation considering other agents' action. The proposed system evaluates office layout plans for polygonal space generated by the office layout support system using genetic algorithm. In the proposed system, the office layout plan is given, and then the agents move under the conditions. Each agent decides the escape route based on the information about office layout, impassable spaces, crowded areas, other agents' action and so on, and goes to the entrance. Based on the behavior of agents, the evaluation on the maximum time for escape, average speed, the number of agents who could not reach entrance and so on are carried out. We carried out a series of computer experiments in order to demonstrate the effectiveness of the proposed system and confirmed that the proposed system can evaluate layout plans.