The present disclosure provides an atrial fibrillation signal recognition method, apparatus and device. The method comprises: obtaining an electrocardiogram signal to be recognized; inputting said electrocardiogram signal to a pre-established atrial fibrillation signal recognition model, and outputting an atrial fibrillation signal recognition result, wherein the atrial fibrillation signal recognition model is established in the following way: obtaining a set number of electrocardiogram sample signals and corresponding identifier information, wherein the identifier information comprises identifier information of normal signals and of atrial fibrillation signals; balancing, according to the number of normal signals, atrial fibrillation signals by means of SMOTE; establishing a network structure of multiple convolutional neural networks, each of said convolutional neural networks being provided with a specific receptive field for recognizing the atrial fibrillation signals of corresponding granularity; inputting the normal signals and the balanced atrial fibrillation signals to the network structure for training to generate an atrial fibrillation signal recognition model. The present disclosure can improve the reliability and accuracy of atrial fibrillation signal recognition.La présente invention concerne un procédé, un appareil et un dispositif de reconnaissance de signaux de fibrillation auriculaire. Le procédé comporte les étapes suivantes : l'obtention d'un signal d'électrocardiogramme à reconnaître ; l'entrée dudit signal d'électrocardiogramme dans un modèle de reconnaissance de signaux de fibrillation auriculaire préétabli, et la sortie d'un résultat de reconnaissance de signaux de fibrillation auriculaire, le modèle de reconnaissance de signaux de fibrillation auriculaire étant établi de la manière suivante : l'obtention d'un nombre d'ensemble de signaux d'échantillon d'électrocardiogramme et d'informations d'identifiant correspondantes, les informations d'ide