Abstract: A method and an associated device for detecting fraud during automatic face recognition, the method comprising the following steps: acquiring a first image of the face by means of a first sensor having a first field angle, and a second image of the face by means of a second sensor having a second field angle that is narrower than the first field angle; analyzing the first image to verify that there is no frame around the face; and analyzing the second image to verify that there is no moiré effect.

Abstract: A method for secure learning of parameters of a convolution neural network, CNN, for data classification includes the implementation, by data processing of a first server, including receiving from a second server a base of already classified learning data, the learning data being homomorphically encrypted; learning in the encrypted domain, from the learning database, the parameters of a reference CNN including a non-linear layer (POLYNOMIAL) operating an at least two-degree polynomial function approximating an activation function; a batch normalization layer before each non-linear layer (POLYNOMIAL); and transmitting the learnt parameters to the second server, for decryption and use for classification.

Abstract: The present invention relates to a PCR kit for improving detection of particular nucleic acid motifs and for increasing the reliability of the results obtained. More precisely, the invention proposes to use modified primers having particular characteristics making it possible, during electrophoresis treatment, to detect, on samples containing nucleic acids, the equivalent of a characteristic encoding of particular nucleic acid motifs, this encoding being such that it makes it possible to better detect the presence of these motifs. The invention also proposes a method for detecting nucleic acid motifs of interest (for example STRs, mini-STRs or VNTRs) using such primers. It advantageously finds—but is not limited to—application in the field of capillary electrophoresis.