Abstract: Micro-expressions are facial muscle movements that reveal a person’s true emotions and are widely-used in fields such as lie detection. It has the characteristics of small action amplitude and short duration, which brings great challenges to micro-expression recognition. To address the above problems, this paper proposes a micro-expression recognition method combining Inception with improved attention mechanism CBAM+. The optical flow information of micro-expression segments was extracted. The Inception module was used to extract the multi-scale features of the face. Parallel channel attention and spatial attention mechanism CBAM+ were proposed to extract features that contribute more to the recognition task. Experiments were performed on the mixed dataset MEGC2019. The unweighted F1-score (UF1) and unweighted average recall (UAR) are 0.7420 and 0.7435, respectively, outperforming the traditional and popular deep learning methods.