Visual inspection is one of the most common non-destructive testing (NDT) methods that offers a fast evaluation of surface damage in aerospace composite structures. However, it is highly dependent on human-related factors and may not detect barely visible impact damage (BVID). In this research, low velocity impact tests with different energy levels are conducted on two groups of composite panels, namely ‘reference’ and ‘sensor-integrated’ samples. Then, the results of impact tests, together with C-scan and visual inspection images, are analysed to define the BVID range and create an original image dataset. Next, four different deep learning models are trained, validated and tested to capture the BVID only from the images of the impacted and non-impacted surfaces. The results show that all four networks can learn and detect BVID quite well, and the sensor-integrated samples reduce the training time and improve the accuracy of deep learning models. ResNet outperforms other networks with the highest accuracy of 96.2% and 98.36% on the back-face of reference and sensor-integrated samples, respectively. The proposed damage recognition method can act as a fast, inexpensive and accurate structural health monitoring tool for composite structures in real-life applications.

Computer-aided detection aims to improve breast cancer screening programs by helping radiologists to evaluate digital mammography (DM) exams. DM exams are generated by devices from different vendors, with diverse characteristics between and even within vendors. Physical properties of these devices and postprocessing of the images can greatly influence the resulting mammogram. This results in the fact that a deep learning model trained on data from one vendor cannot readily be applied to data from another vendor. This paper investigates the use of tailored transfer learning methods based on adversarial learning to tackle this problem. We consider a database of DM exams (mostly bilateral and two views) generated by Hologic and Siemens vendors. We analyze two transfer learning settings: 1) unsupervised transfer, where Hologic data with soft lesion annotation at pixel level and Siemens unlabelled data are used to annotate images in the latter data; 2) weak supervised transfer, where exam level labels for images from the Siemens mammograph are available. We propose tailored variants of recent state-of-the-art methods for transfer learning which take into account the class imbalance and incorporate knowledge provided by the annotations at exam level. Results of experiments indicate the beneficial effect of transfer learning in both transfer settings. Notably, at 0.02 false positives per image, we achieve a sensitivity of 0.37, compared to 0.30 of a baseline with no transfer. Results indicate that using exam level annotations gives an additional increase in sensitivity.

In this paper, we investigate the connection between how people understand speech and how speech is understood by a deep neural network. A naïve, general feed-forward deep neural network was trained for the task of vowel/consonant classification. Subsequently, the representations of the speech signal in the different hidden layers of the DNN were visualized. The visualizations allow us to study the distance between the representations of different types of input frames and observe the clustering structures formed by these representations. In the different visualizations, the input frames were labeled with different linguistic categories: sounds in the same phoneme class, sounds with the same manner of articulation, and sounds with the same place of articulation. We investigate whether the DNN clusters speech representations in a way that corresponds to these linguistic categories and observe evidence that the DNN does indeed appear to learn structures that humans use to understand speech without being explicitly trained to do so.