- Guided Attention Inference Network. [Journal Article]
- ITIEEE Trans Pattern Anal Mach Intell 2019 Jun 07
- With only coarse labels, weakly supervised learning typically uses top-down attention maps generated by back-propagating gradients as priors for tasks such as object localization and semantic segment…
With only coarse labels, weakly supervised learning typically uses top-down attention maps generated by back-propagating gradients as priors for tasks such as object localization and semantic segmentation. While these attention maps are intuitive and informative explanations of deep neural network, there is no effective mechanism to manipulate the network attention during learning process. In this paper, we address three shortcomings of previous approaches in modeling such attention maps in one common framework. First, we make attention maps a natural and explicit component in the training pipeline such that they are end-to-end trainable. Moreover, we provide self-guidance directly on these maps by exploring supervision from the network itself to improve them towards specific target tasks. Lastly, we proposed a design to seamlessly bridge the gap between using weak and extra supervision if available. Despite its simplicity, experiments on the semantic segmentation task demonstrate the effectiveness of our methods. Besides, the proposed framework provides a way not only explaining the focus of the learner but also feeding back with direct guidance towards specific tasks. Under mild assumptions our method can also be understood as a plug-in to existing convolutional neural networks to improve their generalization performance.
- Feature Pyramid Reconfiguration with Consistent Loss for Object Detection. [Journal Article]
- ITIEEE Trans Image Process 2019 May 24
- Taking the feature pyramids into account has become a crucial way to boost the object detection performance. While various pyramid representations have been developed, previous works are still ineffi…
Taking the feature pyramids into account has become a crucial way to boost the object detection performance. While various pyramid representations have been developed, previous works are still inefficient to integrate the semantical information over different scales. Moreover, recent object detectors are suffering from accurate object location applications, mainly due to the coarse definition of the "positive" examples at training and predicting phases. In this paper, we begin by analyzing current pyramid solutions, and then propose a novel architecture by reconfiguring the feature hierarchy in a flexible yet effective way. In particular, our architecture consists of two lightweight and trainable processes: global attention and local reconfiguration. The global attention is to emphasize the global information of each feature scale, while the local reconfiguration is to capture the local correlations across different scales. Both the global attention and local reconfiguration are non-linear and thus exhibit more expressive ability. Then, we discover that the loss function for object detectors during training is the central cause of the inaccurate location problem. We propose to address this issue by reshaping the standard cross entropy loss such that it focuses more on accurate predictions. Both the feature reconfiguration and the consistent loss could be utilized in popular one-stage (SSD, RetinaNet) and two-stage (Faster R-CNN) detection frameworks. Extensive experimental evaluations on PASCAL VOC 2007, PASCAL VOC 2012 and MS COCO datasets demonstrate that, our models achieve consistent and significant boosts compared with other state-of-the-art methods.
- Ultrastructural Imaging Analysis of the Zona Pellucida Surface in Bovine Oocytes. [Journal Article]
- MMMicrosc Microanal 2019 May 28; :1-5
- The aims of the present study were to: (i) evaluate the ultrastructural differences in the zona pellucida (ZP) surface between immature and mature bovine oocytes, and (ii) describe a new objective te…
The aims of the present study were to: (i) evaluate the ultrastructural differences in the zona pellucida (ZP) surface between immature and mature bovine oocytes, and (ii) describe a new objective technique to measure the pores in the outer ZP. Intact cumulus-oocyte complexes (COCs) obtained from a local abattoir were immediately fixed (immature group) or submitted to in vitro maturation (IVM) at 38.5 °C for 24 h in a humidified atmosphere of 5% CO2 in air (mature group). Oocytes from both groups were morphologically evaluated via Scanning Electron Microscopy (SEM) and the images were processed in the Fiji/ImageJ software using a new objective methodology through the Trainable Weka Segmentation plugin. The average number of pores in ZP was greater (p 0.05) between groups. In conclusion, it has been shown that the number of pores highlighted the main ultrastructural change in the morphology of the ZP surface of bovine oocytes during the IVM process. We have described an objective method that can be used to evaluate ultrastructural modifications of the ZP surface during oocyte maturation and early embryo development.
- Computationally Efficient Deep Neural Network for Computed Tomography Image Reconstruction. [Journal Article]
- MPMed Phys 2019 May 27
- CONCLUSIONS: We proposed a training-time computationally efficient neural network for CT image reconstruction. The proposed method achieved comparable image quality with state-of-the-art neural network for CT reconstruction, with significantly reduced memory and time requirement during training. The proposed method is applicable to 3D image reconstruction problems such as cone-beam CT and tomosynthesis on mainstream GPUs. This article is protected by copyright. All rights reserved.
- Core competencies for academic leaders in Iran University of Medical Sciences. [Journal Article]
- AMAdv Med Educ Pract 2019; 10:221-228
- CONCLUSIONS: In order to design a development program for academic leaders in the integrated educational system in Iran University of Medical Sciences, both personal and functional competencies need to be considered together.
- ROSA: Robust Salient Object Detection Against Adversarial Attacks. [Journal Article]
- ITIEEE Trans Cybern 2019 May 17
- Recently, salient object detection has witnessed remarkable improvement owing to the deep convolutional neural networks which can harvest powerful features for images. In particular, the state-of-the…
Recently, salient object detection has witnessed remarkable improvement owing to the deep convolutional neural networks which can harvest powerful features for images. In particular, the state-of-the-art salient object detection methods enjoy high accuracy and efficiency from fully convolutional network (FCN)-based frameworks which are trained from end to end and predict pixel-wise labels. However, such framework suffers from adversarial attacks which confuse neural networks via adding quasi-imperceptible noises to input images without changing the ground truth annotated by human subjects. To our knowledge, this paper is the first one that mounts successful adversarial attacks on salient object detection models and verifies that adversarial samples are effective on a wide range of existing methods. Furthermore, this paper proposes a novel end-to-end trainable framework to enhance the robustness for arbitrary FCN-based salient object detection models against adversarial attacks. The proposed framework adopts a novel idea that first introduces some new generic noise to destroy adversarial perturbations, and then learns to predict saliency maps for input images with the introduced noise. Specifically, our proposed method consists of a segment-wise shielding component, which preserves boundaries and destroys delicate adversarial noise patterns and a context-aware restoration component, which refines saliency maps through global contrast modeling. The experimental results suggest that our proposed framework improves the performance significantly for state-of-the-art models on a series of datasets.
- A Unified Novel Neural Network Approach and a Prototype Hardware Implementation for Ultra-Low Power EEG Classification. [Journal Article]
- ITIEEE Trans Biomed Circuits Syst 2019 May 15
- This paper introduces a novel electroencephalogram (EEG) data classification scheme together with its implementation in hardware using an innovative approach. The proposed scheme integrates into a si…
This paper introduces a novel electroencephalogram (EEG) data classification scheme together with its implementation in hardware using an innovative approach. The proposed scheme integrates into a single, end-to-end trainable model a spatial filtering technique and a neural network based classifier. The spatial filters, as well as, the coefficients of the neural network classifier are simultaneously estimated during training. By using different time-locked spatial filters, we introduce for the first time the notion of "attention" in EEG processing, which allows for the efficient capturing of the temporal dependencies and/ or variability of the EEG sequential data. One of the most important benefits of our approach is that the proposed classifier is able to construct highly discriminative features directly from raw EEG data and, at the same time, to exploit the function approximation properties of neural networks, in order to produce highly accurate classification results. The evaluation of the proposed methodology, using public available EEG datasets, indicates that it outperforms the standard EEG classification approach based on filtering and classification as two separated steps. Moreover, we present a prototype implementation of the proposed scheme in state-of-the-art reconfigurable hardware; our novel implementation outperforms by more than an order of magnitude, in terms of power efficiency, the conventional CPU-based approaches.
- It's all about the sex, or is it? Humans, horses and temperament. [Journal Article]
- PlosPLoS One 2019; 14(5):e0216699
- We propose that the anthropomorphic application of gender stereotypes to animals influences human-animal interactions and human expectations, often with negative consequences for female animals. An o…
We propose that the anthropomorphic application of gender stereotypes to animals influences human-animal interactions and human expectations, often with negative consequences for female animals. An online survey was conducted to explore riders' perceptions of horse temperament and suitability for ridden work, based on horse sex. The questionnaire asked respondents to allocate three hypothetical horses (a mare, gelding and stallion) to four riders compromising a woman, man, girl and boy. Riders were described as equally capable of riding each horse and each horse was described as suitable for all riders. Participants were also asked which horses (mares, geldings or stallions) were most suitable for the three equestrian disciplines of show-jumping, dressage and trail-riding. Logistic regression analyses were conducted to investigate people's perceptions about suitability of horse types for particular riders, to evaluate if age, strength or gender were important in rider choice and to investigate riders' allocation of various descriptors to a gelding, stallion or mare. There were 1,233 survey respondents, 94% of whom were female and 75% of whom were riders with at least eight years of experience. Binomial logistic regression revealed the girl had 2.5 times the odds of being allocated the gelding compared to the boy (p < 0.001). Respondents were significantly more likely to allocate the stallion to the man and nearly 50% of respondents did not allocate a horse to the boy, even though they ranked rider gender as least important to their choice (p < 0.001). In a forced choice selection of a positive or negative descriptor from a series of nine paired terms to describe horse temperament, a greater proportion of respondents assigned geldings positive ratings on terms such as calm, trainable, reliable and predictable. In terms of suitability for the three equestrian disciplines of show-jumping, dressage and trail-riding, participants overwhelmingly chose geldings for trail-riding, with mares being least preferred for both dressage and show-jumping disciplines. The results suggest that female riders are entering the horse-human dyad with gendered ideas about horse temperament and view horse-riding as an activity primarily for women and girls. This could have far-reaching implications for equine training and welfare.
- Is resilience a trainable skill? [Journal Article]
- BMJBMJ 2019 May 13; 365:l2162
New Search Next
- FP2VEC: a new molecular featurizer for learning molecular properties. [Journal Article]
- BBioinformatics 2019 May 09
- One of the most successful methods for predicting the properties of chemical compounds is the quantitative structure-activity relationship (QSAR) methods. The prediction accuracy of QSAR models has r…
One of the most successful methods for predicting the properties of chemical compounds is the quantitative structure-activity relationship (QSAR) methods. The prediction accuracy of QSAR models has recently been greatly improved by employing deep learning technology. Especially, newly developed molecular featurizers based on graph convolution operations on molecular graphs significantly outperform the conventional extended connectivity fingerprints (ECFP) feature in both classification and regression tasks, indicating that it is critical to develop more effective new featurizers to fully realize the power of deep learning techniques. Motivated by the fact that there is a clear analogy between chemical compounds and natural languages, this work develops a new molecular featurizer, FP2VEC, which represents a chemical compound as a set of trainable embedding vectors.