Proceedings of International Joint Conference on Neural Networks 2020 (IJCNN 2020)
2020 International Joint Conference on Neural Networks (IJCNN) held virtually, as part of the IEEE World Congress on Computational Intelligence (IEEE WCCI) 2020. IJCNN 2020 is jointly organized by the IEEE Computational Intelligence Society (CIS) and the International Neural Network Society (INNS). For IJCNN 2020 (and when WCCI is organized in even-numbered years) IEEE CIS is the lead society and financial sponsor. IJCNN 2020 is the major event in the field of neural networks and learning systems, covering all topics in the field from theory to applications. IJCNN provides a forum for researchers, students and professionals in the field of Neural Network and Learning Systems. The meeting is a unique opportunity to present our research to other colleagues and exchange the latest advances in theories, technologies and practices. It is tremendous opportunity also to know what the trending topics are, the current state-of-the-art and the main applications of Neural Networks and Learning Systems. IJCNN 2020 accepted 1134 papers for inclusion in the conference program at an acceptance rate of 57%. Out of this, 608 papers are being presented in oral sessions and 526 in poster sessions. The largest contributors by country are China (29.7%), USA (15.7%), UK (15.2%), Brazil (10.1%), Australia (8.8%), Japan (7.8%) and India (7.1%). The country assigned to a paper was the country from which its first author came. The program of IJCNN 2020 reflects a rich variety of topics: Deep Learning, Extreme Learning Machines, Feed forward NNs and Supervised Learning, Online and Incremental Learning, Spiking Neural Networks, Unsupervised Learning and Clustering, ADP and Reinforcement Learning, Recurrent NNs and Reservoir Networks, Concept Drift, ML Methods Robust to Large Outliers, Complex Valued NNs, Neural Models and Computation, Memory and Sensory Systems, Semi-supervised Learning and Neuromorphic Hardware. Likewise, a large number of papers deal with a great variety of applications.
In this paper the problem of high computational complexity of deep convolutional nets in image recognition is considered. An existing framework of adaptive neural networks is extended by appending the separate classifier to intermediate layers. The hierarchical representations of the input image are sequentially analyzed. If the first classifier returns rather high confidence score, the inference process will be terminated. Otherwise, the inference to the next intermediate layer with attached classifier is continued until the reliable solution is obtained or the penultimate layer is reached. The thresholds for classifier scores at each layer are automatically chosen based on the Benjamini-Hochberg multiple comparisons for a specified confidence level. Experimental study for both pre-trained and fine-tuned deep convolutional neural networks demonstrates that the proposed approach reduces the running time by up to 1.7 times without significant accuracy degradation. Moreover, the larger is the training sample, the more noticeable is the gain in performance.
In this paper a new formulation of event recognition task is examined: it is required to predict event categories given a gallery of images, for which albums (groups of photos corresponding to a single event) are unknown. The novel two-stage approach is proposed. At first, features are extracted in each photo using the pre-trained convolutional neural network (CNN). These features are classified individually. The normalized scores of the classifier are used to group sequential photos into several clusters. Finally, the features of photos in each group are aggregated into a single descriptor using neural attention mechanism. This algorithm is implemented in Android mobile application. Experimental study with features extracted by contemporary convolutional neural networks including EfficientNets for Photo Event Collection and Multi-Label Curation of Flickr Events Dataset demonstrates that the proposed approach is 9-23% more accurate than conventional event recognition on single photos. Moreover, proposed method has 13-16% lower error rate when compared to classification of groups of photos obtained with hierarchical clustering of CNN-based embeddings.