The partial discharge (PD) phenomenon is one of the major factors that can lead to insulation deterioration in power transformers. Therefore, continuous monitoring of the insulation status and the detection of PD activities may assist in correctly evaluating the insulation, with any required actions being taken accordingly. Acoustic detection has been applied for detecting and locating PD activities inside power transformers. Although the acoustic detection is immune to electromagnetic interference, the acoustic signals suffer from high attenuation, which makes the detection of PD activities a difficult task. This paper presents a pattern recognition-based technique for enhancing the acoustic detection of partial discharge signals. Different cases for PD generation were simulated, which included the presence of different types of barriers such as cellulose insulation paper and silicon steel core material. In addition, the effects of the tank size and the distance between the PD source and the acoustic sensor on the detection performance were studied. The features extracted from the acquired signals in all cases were fed to an artificial neural network, which was used for training and classification. The results show that the detection performance of acoustic PD signals could be significantly enhanced using features such as signal entropy.