Search

“Everything depends on our capacity as humanity to trace an ethical path, to define the ways of useful and reasonable uses” (D. Gruson). Artificial iIntelligence (AI), as evoked by the term, aims to reproduce human thought, and most of the field of robotics is based on a certain level of AI. Using IA in different areas of everyday life is a very active sector in Romania. The development of artificial intelligence (AI) in health is currently experiencing a very strong acceleration with the broad multiplication of applications, particularly in the field of image recognition in radiology, ophthalmology or dermatology. Algorithmic medicine has already become a reality and will become increasingly important in the years to come. Despite the potential benefits of AI in the field of health, there are a number of elements which should be taken into account and which should be integrated because these innovations to be established in society. As no society can anymore function without rules and regulations, this field as well should be regulated (positive regulation). David Gruson (La Machine, le Médecin et Moi, Editions Beta Publisher, 2018) argues in favor of legal regulation of using robots and AI in health. The five keys of regulation for the deployment of artificial intelligence and robotics in health are listed by the author, as following: (1) Information and Patient’s Consent; (2) Guaranteeing the human in AI; (3) Graduation of regulation according to the level of sensitivity of health data; (4) Supporting the adaptation of jobs; (5) Intervention of independent external supervision. These key principles could give the main orientation on the deployment of the AI ethically, in order to separate the societal benefits of societal harms. The AI should offer health care professionals and patients options and better results, and the science should be in the service of humans and of the good of society.

International audience; This paper presents a local no-reference blur assessment method in natural macro-like images. The purpose is to decide the blurriness of the object of interest. In our case, it represents the first step for a plant recognition system. Blur detection works on small non-overlapping blocks using wavelet decomposition and edge classification. At the block level the number of edges is less than on global images. A new set of rules is obtained by a supervised decision tree algorithm trained on a manually labelled base of 1500 blurred/un-blurred images. Our purpose is to achieve a qualitative decision of the blurriness/sharpness of the object of interest making it the first step towards a segmentation process. Experimental results show this method outperforms two other methods found in literature, even if applied on a block basis. Together with a pre-segmentation step, the method allows to decide if the object of interest (leaf, flower) is sharp in order to extract precise botanical key identification features (e. g. leaf border).