Search

Differentiable architecture search (DARTS) transforms architectural optimization into a super network optimization by stacking two cells (2 c.). However, repeatedly stacking two cells is a suboptimal operation since cells in different depths should be various. Besides, we find that the performance is slightly improved by increasing the number of searched cells (e.g., from 2 c. to 5 c.), but it leads to uneven resource allocation. This letter proposes a regularized DARTS (RDARTS) to adjust the architectural differences and balance degrees of freedom and resource allocation. Specifically, we use separate architectural parameters for two reduction cells and three normal cells, and then propose an Reg distance to calculate the difference between cells. We design a new validation loss which is the weighting of cross-entropy and Reg loss and introduce an adaptive adjustment method. Results show that RDARTS achieves the top-1 accuracy of 97.64% and 75.8% on CIFAR and ImageNet. [ABSTRACT FROM AUTHOR]

Porosity produced in laser metal deposition hampers its application due to the absence of an effective prediction method. Measured thermal images of the melt pool provide a unique opportunity for porosity analytics. Furthermore, a physical model may provide complementary rich data that cannot be measured otherwise. How to leverage both types of data to predict porosity is very challenging. This paper presents a physics-driven deep learning model to predict porosity by integrating both measured and predicted data of the melt pool. The model fidelity is validated with the predicted pore occurrence and size with enhanced interpretability of Ti–6Al–4V thin-wall structures. [ABSTRACT FROM AUTHOR]

Emotion recognition is challenging due to the emotional gap between emotions and audio–visual features. Motivated by the powerful feature learning ability of deep neural networks, this paper proposes to bridge the emotional gap by using a hybrid deep model, which first produces audio–visual segment features with Convolutional Neural Networks (CNNs) and 3D-CNN, then fuses audio–visual segment features in a Deep Belief Networks (DBNs). The proposed method is trained in two stages. First, CNN and 3D-CNN models pre-trained on corresponding large-scale image and video classification tasks are fine-tuned on emotion recognition tasks to learn audio and visual segment features, respectively. Second, the outputs of CNN and 3D-CNN models are combined into a fusion network built with a DBN model. The fusion network is trained to jointly learn a discriminative audio–visual segment feature representation. After average-pooling segment features learned by DBN to form a fixed-length global video feature, a linear Support Vector Machine is used for video emotion classification. Experimental results on three public audio–visual emotional databases, including the acted RML database, the acted eNTERFACE05 database, and the spontaneous BAUM-1s database, demonstrate the promising performance of the proposed method. To the best of our knowledge, this is an early work fusing audio and visual cues with CNN, 3D-CNN, and DBN for audio–visual emotion recognition. [ABSTRACT FROM AUTHOR]

Traditional Bag-of-Words (BoWs) model is commonly generated with many steps, including local feature extraction, codebook generation and feature quantization, etc. Those steps are relatively independent with each other and are hard to be jointly optimized. Moreover, the dependency on hand-crafted local feature makes BoWs model not effective in conveying high-level semantics. These issues largely hinder the performance of BoWs model in large-scale image applications. To conquer these issues, we propose an End-to-End BoWs (E2BoWs) model based on Deep Convolutional Neural Network (DCNN). Our model takes an image as input, then identifies and separates semantic objects in it, and finally outputs visual words with high semantic discriminative power. Specifically, our model firstly generates Semantic Feature Maps (SFMs) corresponding to different object categories through convolutional layers, then introduces Bag-of-Words Layers (BoWL) to generate visual words from each individual feature map. We also introduce a novel learning algorithm to reinforce the sparsity of the generated E2BoWs model, which further ensures the time and memory efficiency. We evaluate the proposed E2BoWs model on several image search datasets including MNIST, SVHN, CIFAR-10, CIFAR-100, MIRFLICKR-25K and NUS-WIDE. Experimental results show that our method achieves promising accuracy and efficiency compared with recent deep learning based retrieval works. [ABSTRACT FROM AUTHOR]