This paper presents a novel unsupervised domain adaptation framework, called Synergistic Image and Feature Adaptation (SIFA), to effectively tackle the problem of domain shift. Domain adaptation has become an important and hot topic in recent studies on deep learning, aiming to recover performance degradation when applying the neural networks to new testing domains. Our proposed SIFA is an elegant learning diagram which presents synergistic fusion of adaptations from both image and feature perspectives. In particular, we simultaneously transform the appearance of images across domains and enhance domain-invariance of the extracted features towards the segmentation task. The feature encoder layers are shared by both perspectives to grasp their mutual benefits during the end-to-end learning procedure. Without using any annotation from the target domain, the learning of our unified model is guided by adversarial losses, with multiple discriminators employed from various aspects. We have extensively validated our method with a challenging application of crossmodality medical image segmentation of cardiac structures. Experimental results demonstrate that our SIFA model recovers the degraded performance from 17.2% to 73.0%, and outperforms the state-of-the-art methods by a significant margin.

No abstract

With the ubiquitous deployment of wireless systems and pervasive availability of smart devices, indoor localization is empowering numerous location-based services. With the established radio maps, WiFi fingerprinting has become one of the most accessible and practical approaches to localize a mobile user. However, most fingerprint-based localization algorithms are computation-intensive, with heavy dependence on both offline training phase and online localization phase. In this paper, we propose CNNLoc, a Convolutional Neural Network (CNN) based indoor localization framework with WiFi fingerprints for multi-building and multi-floor localization. We propose a novel classification model by combining Stacked Auto-Encoder (SAE) with one-dimensional CNN. The SAE can be used to extract key features more precisely from sparse Received Signal Strength (RSS) data, and the CNN can be trained to effectively achieve high success rates in the localization phase. We evaluate CNNLoc with state-of-the-arts as benchmarks on the UJIIndoor-Loc dataset and Tampere dataset. CNNLoc shows its excellence in both building-level and floor-level classifications and outperforms the existing solutions with 100% success on building success rate and an average success rate over 95% on floor-level localization.

No abstract

Introduction: We describe the creation of a deep neural network architecture to classify cardiac abnormality from 12 lead ECGs. The model was created by the team "between a ROC and a heart place" for the Phys-ioNet/Computing in Cardiology Challenge 2020.Methods: ECGs were downsampled to 257 Hz and then set to a consistent duration by randomly clipping or zeropadding the signal to 4096 samples. To learn effective features, we created a modified ResNet with larger kernel sizes that models long-term dependencies. We embedded a Squeeze-And-Excitation layer into the modified ResNet to learn the importance of each lead, adaptively. A simple constrained grid-search was applied to deal with class imbalance.Results: Using the bespoke weighted accuracy metric, we achieved a 5-fold cross-validation score of 0.684, sensitivity and specificity of 0.758 and 0.969, respectively. We scored 0.520 on the full test data, and ranked 2nd out of the 41 teams that participated in the challenge.Conclusion: The proposed prediction model performed well on the validation and hidden test data. Such models may be potentially used for ECG screening or diagnosis.