%0 Conference Paper %1 e077c6a8499a4a48a762e4ee1a01e597 %A Vargis, Tom Richard %A Ghiasvand, Siavash %B 2024 29th International Conference on Automation and Computing (ICAC) %C United States of America %D 2024 %I Institute of Electrical and Electronics Engineers Inc. %K livinglab AI, Content-aware Explainable FIS_scads Layered Modular Open adaptive editing, pipeline, restoration, scene source %R 10.1109/ICAC61394.2024.10718754 %T Content-Aware Depth-Adaptive Image Restoration %U https://cacsuk.co.uk/icac/ %X The work prioritizes the development of a modular pipeline that efficiently utilizes existing models for image restoration instead of creating new ones from scratch. Restoration is conducted at an object-specific level, with each object being regenerated based on its corresponding class label information. What sets this approach apart is its provision of comprehensive user control throughout the restoration process. Users have the flexibility to choose models for specific restoration tasks, customize the sequence of steps according to their requirements, and enhance the resulting regenerated image with depth awareness. The study offers two distinct pathways for implementing image regeneration, allowing for a comparative analysis of their strengths and limitations. One of the main advantages of this adaptable system is its flexibility. Users can tailor the restoration process to target specific object categories, such as medical images, by leveraging models trained on those specific object classes. The entire code to replicate the pipeline is publicly accessible on 1. %@ 979-8-3503-6089-9

@inproceedings{e077c6a8499a4a48a762e4ee1a01e597, abstract = {The work prioritizes the development of a modular pipeline that efficiently utilizes existing models for image restoration instead of creating new ones from scratch. Restoration is conducted at an object-specific level, with each object being regenerated based on its corresponding class label information. What sets this approach apart is its provision of comprehensive user control throughout the restoration process. Users have the flexibility to choose models for specific restoration tasks, customize the sequence of steps according to their requirements, and enhance the resulting regenerated image with depth awareness. The study offers two distinct pathways for implementing image regeneration, allowing for a comparative analysis of their strengths and limitations. One of the main advantages of this adaptable system is its flexibility. Users can tailor the restoration process to target specific object categories, such as medical images, by leveraging models trained on those specific object classes. The entire code to replicate the pipeline is publicly accessible on [1].}, added-at = {2024-11-28T16:27:18.000+0100}, address = {United States of America}, author = {Vargis, {Tom Richard} and Ghiasvand, Siavash}, biburl = {https://puma.scadsai.uni-leipzig.de/bibtex/27298f43ff024128aa704a8219409be9e/scadsfct}, booktitle = {2024 29th International Conference on Automation and Computing (ICAC)}, doi = {10.1109/ICAC61394.2024.10718754}, interhash = {31c6825478ba647a8de85f205e0649cf}, intrahash = {7298f43ff024128aa704a8219409be9e}, isbn = {979-8-3503-6089-9}, keywords = {livinglab AI, Content-aware Explainable FIS_scads Layered Modular Open adaptive editing, pipeline, restoration, scene source}, language = {English}, note = {Publisher Copyright: {\textcopyright} 2024 IEEE.; 29th International Conference on Automation and Computing : Smart Systems and Digital Healthcare, ICAC 2024 ; Conference date: 28-08-2024 Through 30-08-2024}, publisher = {Institute of Electrical and Electronics Engineers Inc.}, timestamp = {2024-11-28T17:40:57.000+0100}, title = {Content-Aware Depth-Adaptive Image Restoration}, url = {https://cacsuk.co.uk/icac/}, year = 2024 }