Diabetic retinopathy (DR) is one of the most common eye conditions among diabetic patients. However, vision loss occurs primarily in the late stages of DR, and the symptoms of visual impairment, ranging from mild to severe, can vary greatly, adding to the burden of diagnosis and treatment in clinical practice. Deep learning methods based on retinal images have achieved remarkable success in automatic DR grading, but most of them neglect that the presence of diabetes usually affects both eyes, and ophthalmologists usually compare both eyes concurrently for DR diagnosis, leaving correlations between left and right eyes unexploited. In this study, simulating the diagnostic process, we propose a two-stream binocular network to capture the subtle correlations between left and right eyes, in which, paired images of eyes are fed into two identical subnetworks separately during training. We design a contrastive grading loss to learn binocular correlation for five-class DR detection, which maximizes inter-class dissimilarity while minimizing the intra-class difference. Experimental results on the EyePACS dataset show the superiority of the proposed binocular model, outperforming monocular methods by a large margin.

Accurate automatic liver and tumor segmentation plays a vital role in treatment planning and disease monitoring. Recently, deep convolutional neural network (DCNNs) has obtained tremendous success in 2D and 3D medical image segmentation. However, 2D DCNNs cannot fully leverage the inter-slice information, while 3D DCNNs are computationally expensive and memory intensive. To address these issues, we first propose a novel dense-sparse training flow from a data perspective, in which, densely adjacent slices and sparsely adjacent slices are extracted as inputs for regularizing DCNNs, thereby improving the model performance. Moreover, we design a 2.5D light-weight nnU-Net from a network perspective, in which, depthwise separable convolutions are adopted to improve the efficiency. Extensive experiments on the LiTS dataset have demonstrated the superiority of the proposed method.

To quantitatively and intuitively explore the generalization ability of pre-trained language models (PLMs), we have designed several tasks of arithmetic and logical reasoning. We both analyse how well PLMs generalize when the test data is in the same distribution as the train data and when it is different, for the latter analysis, we have also designed a cross-distribution test set other than the in-distribution test set. We conduct experiments on one of the most advanced and publicly released generative PLM - BART. Our research finds that the PLMs can easily generalize when the distribution is the same, however, it is still difficult for them to generalize out of the distribution.

Zero-shot classification is a promising paradigm to solve an applicable problem when the training classes and test classes are disjoint. Achieving this usually needs experts to externalize their domain knowledge by manually specifying a class-attribute matrix to define which classes have which attributes. Designing a suitable class-attribute matrix is the key to the subsequent procedure, but this design process is tedious and trial-and-error with no guidance. This paper proposes a visual explainable active learning approach with its design and implementation called semantic navigator to solve the above problems. This approach promotes human-AI teaming with four actions (ask, explain, recommend, respond) in each interaction loop. The machine asks contrastive questions to guide humans in the thinking process of attributes. A novel visualization called semantic map explains the current status of the machine. Therefore analysts can better understand why the machine misclassifies objects. Moreover, the machine recommends the labels of classes for each attribute to ease the labeling burden. Finally, humans can steer the model by modifying the labels interactively, and the machine adjusts its recommendations. The visual explainable active learning approach improves humans' efficiency of building zero-shot classification models interactively, compared with the method without guidance. We justify our results with user studies using the standard benchmarks for zero-shot classification.

Generative neural network architectures such as GANs, may be used to generate synthetic instances to compensate for the lack of real data. However, they may be employed to create media that may cause social, political or economical upheaval. One emerging media is "Deepfake". Techniques that can discriminate between such media is indispensable. In this paper, we propose a modified multilinear (tensor) method, a combination of linear and multilinear regressions for representing fake and real data. We test our approach Figure 1: Deepfake technique replaces a person's appearance in by representing Deepfakes with our modified multilinear (tensor) an existing image or video with someone else's appearance [20].

To this end, we propose a novel deep neural network called as Research literature relevant to our work consists of the work done Deep Geospatial Interpolation Network(DGIN), which incorporates in the areas of traditional spatial statistics [1, 7, 9, 10], spatial data both spatial and temporal relationships and has significantly lower mining [4, 8], neural networks for spatio-temporal data [3, 11], and training time. DGIN consists of three major components: Spatial computer vision [5, 6]. Encoder to capture the spatial dependencies, Sequential module Spatial statistics techniques such as IDW [9], DDW [10], Kriging to incorporate the temporal dynamics, and an Attention block to [1, 7], and its variants are not suitable for the interpolation learn the importance of the temporal neighborhood around the problem because of the following reasons: (a) high execution time gap. We evaluate DGIN on the MODIS reflectance dataset from (in case of Kriging), (b) strong assumptions on the nature of spatial two different regions. Our experimental results indicate that DGIN relationships (such as inverse relationship in case of IDW), (c) has two advantages: (a) it outperforms alternative approaches (has prior assumption and/or knowledge on statistical properties of data lower MSE with p-value 0.01) and, (b) it has significantly low (e.g., precise knowledge of the mean in case of Simple Kriging and execution time than Kriging.

Handwritten character recognition (HCR) is a challenging learning problem in pattern recognition, mainly due to similarity in structure of characters, different handwriting styles, noisy datasets and a large variety of languages and scripts. HCR problem is studied extensively for a few decades but there is very limited research on script independent models. This is because of factors, like, diversity of scripts, focus of the most of conventional research efforts on handcrafted feature extraction techniques which are language/script specific and are not always available, and unavailability of public datasets and codes to reproduce the results. On the other hand, deep learning has witnessed huge success in different areas of pattern recognition, including HCR, and provides end-to-end learning, i.e., automated feature extraction and recognition. In this paper, we have proposed a novel deep learning architecture which exploits transfer learning and image-augmentation for end-to-end learning for script independent handwritten character recognition, called HCR-Net. The network is based on a novel transfer learning approach for HCR, where some of lower layers of a pre-trained VGG16 network are utilised. Due to transfer learning and image-augmentation, HCR-Net provides faster training, better performance and better generalisations. The experimental results on publicly available datasets of Bangla, Punjabi, Hindi, English, Swedish, Urdu, Farsi, Tibetan, Kannada, Malayalam, Telugu, Marathi, Nepali and Arabic languages prove the efficacy of HCR-Net and establishes several new benchmarks. For reproducibility of the results and for the advancements of the HCR research, complete code is publicly released at \href{https://github.com/jmdvinodjmd/HCR-Net}{GitHub}.

We introduce a noisy channel approach for language model prompting in few-shot text classification. Instead of computing the likelihood of the label given the input (referred as direct models), channel models compute the conditional probability of the input given the label, and are thereby required to explain every word in the input. We use channel models for recently proposed few-shot learning methods with no or very limited updates to the language model parameters, via either in-context demonstration or prompt tuning. Our experiments show that, for both methods, channel models significantly outperform their direct counterparts, which we attribute to their stability, i.e., lower variance and higher worst-case accuracy. We also present extensive ablations that provide recommendations for when to use channel prompt tuning instead of other competitive models (e.g., direct head tuning): channel prompt tuning is preferred when the number of training examples is small, labels in the training data are imbalanced, or generalization to unseen labels is required.

Accurate prediction of short-term OD Matrix (i.e. the distribution of passenger flows from various origins to destinations) is a crucial task in metro systems. It is highly challenging due to the constantly changing nature of many impacting factors and the real-time de- layed data collection problem. Recently, some deep learning-based models have been proposed for OD Matrix forecasting in ride- hailing and high way traffic scenarios. However, these models can not sufficiently capture the complex spatiotemporal correlation between stations in metro networks due to their different prior knowledge and contextual settings. In this paper we propose a hy- brid framework Multi-view TRGRU to address OD metro matrix prediction. In particular, it uses three modules to model three flow change patterns: recent trend, daily trend, weekly trend. In each module, a multi-view representation based on embedding for each station is constructed and fed into a transformer based gated re- current structure so as to capture the dynamic spatial dependency in OD flows of different stations by a global self-attention mecha- nism. Extensive experiments on three large-scale, real-world metro datasets demonstrate the superiority of our Multi-view TRGRU over other competitors.

Accurate traffic state prediction is the foundation of transportation control and guidance. It is very challenging due to the complex spatiotemporal dependencies in traffic data. Existing works cannot perform well for multi-step traffic prediction that involves long future time period. The spatiotemporal information dilution becomes serve when the time gap between input step and predicted step is large, especially when traffic data is not sufficient or noisy. To address this issue, we propose a multi-spatial graph convolution based Seq2Seq model. Our main novelties are three aspects: (1) We enrich the spatiotemporal information of model inputs by fusing multi-view features (time, location and traffic states) (2) We build multiple kinds of spatial correlations based on both prior knowledge and data-driven knowledge to improve model performance especially in insufficient or noisy data cases. (3) A spatiotemporal attention mechanism based on reachability knowledge is novelly designed to produce high-level features fed into decoder of Seq2Seq directly to ease information dilution. Our model is evaluated on two real world traffic datasets and achieves better performance than other competitors.