ESR14 - Kooshan Hashemifard

ESR14 - Kooshan Hashemifard
Research project
Context recognition for the application of visual privacy
About the project

Most works conceal people’s visual privacy by using blurring or pixelating effects to modify an image. In a privacy-by-context approach, a level-based visualisation scheme to protect privacy is proposed. Each level establishes the way in which the video images are modified and displayed and, therefore, the provided protection degree. In this scheme, the appropriate level is dynamically selected according to the context, therefore modifying a non-protected image before it is displayed. The context has to provide enough information in order to empower monitored people to adapt privacy to their preferences, in such a way that they can decide by whom, how and when they can be watched. The context is modelled by different variables: (i) the identity of the observer; (ii) the identity of the person (to retrieve the privacy profile); (iii) the closeness between the person and observer (e.g., relative, doctor or acquaintance); (iv) appearance (dressed?); (v) location (e.g., kitchen); and (vi) ongoing activity or detected event (e.g., cooking, watching TV, fall). Therefore, an accurate recognition of the context is paramount to provide the appropriate privacy level. This project has investigated techniques to recognise accurately these variables and it has validated them under different use case scenarios.

Start date: April 2021

Expected end date: Autumn 2025

Progress of the project

ESR14 focused on enhancing the effectiveness and privacy preservation of video-based AAL technologies through the integration of context-aware components. ESR14 has worked primarily in three key privacy variables—nudity detection, accident detection, and activity recognition—that help dynamically adapt system behaviour to balance user privacy and safety. A level-based nudity detection system anonymises sensitive visuals depending on exposure levels, while advanced fall detection algorithms help identify emergencies like falls without invading privacy. Daily activity recognition enables the system to adjust monitoring according to the user’s actions, such as resting, cooking, or exercising. These combined capabilities ensure that privacy is maintained without compromising essential care functions, improving user trust and acceptance.


To support these goals, several technical contributions were made. A large-scale Human Skin Segmentation Dataset was developed to improve nudity detection, addressing challenges like varying skin tones and lighting. A skin segmentation model using attention modules was trained to enhance robustness and accuracy. A fallen person detection system using a mobile robot and edge-AI cameras demonstrated high precision and real-time performance. Additionally, a daily activity recognition system, built on 3D CNNs and transformers, enabled accurate classification of routine and emergency behaviours. By integrating these systems, the study advances the implementation of Privacy-by-Context—allowing AAL technologies to adapt privacy controls based on activity, time, and user preferences—bringing context-aware, privacy-respecting video monitoring closer to real-world deployment.

Scientific publications
About the ESR

Kooshan obtained his bachelor degree in Electrical Engineering (2015) from KNTU and a master’s degree focused on Signal Processing (2018) from Iran Broadcasting University. During his studies, he mainly explored the field of machine learning and after graduation, he worked as a Computer Vision engineer in different startups in Iran and developed large-scale machine learning services.

Contact information

Kooshan Hashemifard

University of Alicante
Department of Computing Technology

Ctra. San Vicente del Raspeig, S/N
03690 San Vicente del Raspeig, Spain

Email address: k.hashemifard@ua.es

Partners