Temporal action segmentation approaches have been very successful recently. However, annotating videos with frame-wise labels to train such models is very expensive and time consuming. While weakly supervised methods trained using only ordered action lists require much less annotation effort, the performance is still much worse than fully supervised approaches. In this paper, we introduce timestamp supervision for the temporal action segmentation task. Timestamps require a comparable annotation effort to weakly supervised approaches, and yet provide a more supervisory signal. To demonstrate the effectiveness of timestamp supervision, we propose an approach to train a segmentation model using only timestamps annotations. Our approach uses the model output and the annotated timestamps to generate frame-wise labels by detecting the action changes. We further introduce a confidence loss that forces the predicted probabilities to monotonically decrease as the distance to the timestamps increases. This ensures that all and not only the most distinctive frames of an action are learned during training. The evaluation on four datasets shows that models trained with timestamps annotations achieve comparable performance to the fully supervised approaches.
Handwritten mathematical expression recognition (HMER) is an important research direction in handwriting recognition. The performance of HMER suffers from the two-dimensional structure of mathematical expressions (MEs). To address this issue, in this paper, we propose a high-performance HMER model with scale augmentation and drop attention. Specifically, tackling ME with unstable scale in both horizontal and vertical directions, scale augmentation improves the performance of the model on MEs of various scales. An attention-based encoder-decoder network is used for extracting features and generating predictions. In addition, drop attention is proposed to further improve performance when the attention distribution of the decoder is not precise. Compared with previous methods, our method achieves state-of-the-art performance on two public datasets of CROHME 2014 and CROHME 2016.
We present joint multi-dimension pruning (named as JointPruning), a new perspective of pruning a network on three crucial aspects: spatial, depth and channel simultaneously. The joint strategy enables to search a better status than previous studies that focused on individual dimension solely, as our method is optimized collaboratively across the three dimensions in a single end-to-end training. Moreover, each dimension that we consider can promote to get better performance through colluding with the other two. Our method is realized by the adapted stochastic gradient estimation. Extensive experiments on large-scale ImageNet dataset across a variety of network architectures MobileNet V1&V2 and ResNet demonstrate the effectiveness of our proposed method. For instance, we achieve significant margins of 2.5% and 2.6% improvement over the state-of-the-art approach on the already compact MobileNet V1&V2 under an extremely large compression ratio.
In this paper, we propose an efficient method for robust 3D self-portraits using a single RGBD camera. Benefiting from the proposed PIFusion and lightweight bundle adjustment algorithm, our method can generate detailed 3D self-portraits in seconds and shows the ability to handle subjects wearing extremely loose clothes. To achieve highly efficient and robust reconstruction, we propose PIFusion, which combines learning-based 3D recovery with volumetric non-rigid fusion to generate accurate sparse partial scans of the subject. Moreover, a non-rigid volumetric deformation method is proposed to continuously refine the learned shape prior. Finally, a lightweight bundle adjustment algorithm is proposed to guarantee that all the partial scans can not only "loop" with each other but also remain consistent with the selected live key observations. The results and experiments show that the proposed method achieves more robust and efficient 3D self-portraits compared with state-of-the-art methods.
Outlier detection is an important task in data mining and many technologies have been explored in various applications. However, due to the default assumption that outliers are non-concentrated, unsupervised outlier detection may not correctly detect group anomalies with higher density levels. As for the supervised outlier detection, although high detection rates and optimal parameters can usually be achieved, obtaining sufficient and correct labels is a time-consuming task. To address these issues, we focus on semi-supervised outlier detection with few identified anomalies, in the hope of using limited labels to achieve high detection accuracy. First, we propose a novel detection model Dual-GAN, which can directly utilize the potential information in identified anomalies to detect discrete outliers and partially identified group anomalies simultaneously. And then, considering the instances with similar output values may not all be similar in a complex data structure, we replace the two MO-GAN components in Dual-GAN with the combination of RCC and M-GAN (RCC-Dual-GAN). In addition, to deal with the evaluation of Nash equilibrium and the selection of optimal model, two evaluation indicators are created and introduced into the two models to make the detection process more intelligent. Extensive experiments on both benchmark datasets and two practical tasks demonstrate that our proposed approaches (i.e., Dual-GAN and RCC-Dual-GAN) can significantly improve the accuracy of outlier detection even with only a few identified anomalies. Moreover, compared with the two MO-GAN components in Dual-GAN, the network structure combining RCC and M-GAN has greater stability in various situations.
In the past decade, there has been substantial progress at training increasingly deep neural networks. Recent advances within the teacher--student training paradigm have established that information about past training updates show promise as a source of guidance during subsequent training steps. Based on this notion, in this paper, we propose Long Short-Term Sample Distillation, a novel training policy that simultaneously leverages multiple phases of the previous training process to guide the later training updates to a neural network, while efficiently proceeding in just one single generation pass. With Long Short-Term Sample Distillation, the supervision signal for each sample is decomposed into two parts: a long-term signal and a short-term one. The long-term teacher draws on snapshots from several epochs ago in order to provide steadfast guidance and to guarantee teacher--student differences, while the short-term one yields more up-to-date cues with the goal of enabling higher-quality updates. Moreover, the teachers for each sample are unique, such that, overall, the model learns from a very diverse set of teachers. Comprehensive experimental results across a range of vision and NLP tasks demonstrate the effectiveness of this new training method.
Despite impressive performance on numerous visual tasks, Convolutional Neural Networks (CNNs) --- unlike brains --- are often highly sensitive to small perturbations of their input, e.g. adversarial noise leading to erroneous decisions. We propose to regularize CNNs using large-scale neuroscience data to learn more robust neural features in terms of representational similarity. We presented natural images to mice and measured the responses of thousands of neurons from cortical visual areas. Next, we denoised the notoriously variable neural activity using strong predictive models trained on this large corpus of responses from the mouse visual system, and calculated the representational similarity for millions of pairs of images from the model's predictions. We then used the neural representation similarity to regularize CNNs trained on image classification by penalizing intermediate representations that deviated from neural ones. This preserved performance of baseline models when classifying images under standard benchmarks, while maintaining substantially higher performance compared to baseline or control models when classifying noisy images. Moreover, the models regularized with cortical representations also improved model robustness in terms of adversarial attacks. This demonstrates that regularizing with neural data can be an effective tool to create an inductive bias towards more robust inference.
Accurate multi-organ abdominal CT segmentation is essential to many clinical applications such as computer-aided intervention. As data annotation requires massive human labor from experienced radiologists, it is common that training data are partially labeled, e.g., pancreas datasets only have the pancreas labeled while leaving the rest marked as background. However, these background labels can be misleading in multi-organ segmentation since the "background" usually contains some other organs of interest. To address the background ambiguity in these partially-labeled datasets, we propose Prior-aware Neural Network (PaNN) via explicitly incorporating anatomical priors on abdominal organ sizes, guiding the training process with domain-specific knowledge. More specifically, PaNN assumes that the average organ size distributions in the abdomen should approximate their empirical distributions, a prior statistics obtained from the fully-labeled dataset. As our training objective is difficult to be directly optimized using stochastic gradient descent [20], we propose to reformulate it in a min-max form and optimize it via the stochastic primal-dual gradient algorithm. PaNN achieves state-of-the-art performance on the MICCAI2015 challenge "Multi-Atlas Labeling Beyond the Cranial Vault", a competition on organ segmentation in the abdomen. We report an average Dice score of 84.97%, surpassing the prior art by a large margin of 3.27%.
Deep neural networks (DNNs), especially deep convolutional neural networks (CNNs), have emerged as the powerful technique in various machine learning applications. However, the large model sizes of DNNs yield high demands on computation resource and weight storage, thereby limiting the practical deployment of DNNs. To overcome these limitations, this paper proposes to impose the circulant structure to the construction of convolutional layers, and hence leads to circulant convolutional layers (CircConvs) and circulant CNNs. The circulant structure and models can be either trained from scratch or re-trained from a pre-trained non-circulant model, thereby making it very flexible for different training environments. Through extensive experiments, such strong structure-imposing approach is proved to be able to substantially reduce the number of parameters of convolutional layers and enable significant saving of computational cost by using fast multiplication of the circulant tensor.
Recurrent Neural Networks (RNNs) are becoming increasingly important for time series-related applications which require efficient and real-time implementations. The two major types are Long Short-Term Memory (LSTM) and Gated Recurrent Unit (GRU) networks. It is a challenging task to have real-time, efficient, and accurate hardware RNN implementations because of the high sensitivity to imprecision accumulation and the requirement of special activation function implementations. A key limitation of the prior works is the lack of a systematic design optimization framework of RNN model and hardware implementations, especially when the block size (or compression ratio) should be jointly optimized with RNN type, layer size, etc. In this paper, we adopt the block-circulant matrix-based framework, and present the Efficient RNN (E-RNN) framework for FPGA implementations of the Automatic Speech Recognition (ASR) application. The overall goal is to improve performance/energy efficiency under accuracy requirement. We use the alternating direction method of multipliers (ADMM) technique for more accurate block-circulant training, and present two design explorations providing guidance on block size and reducing RNN training trials. Based on the two observations, we decompose E-RNN in two phases: Phase I on determining RNN model to reduce computation and storage subject to accuracy requirement, and Phase II on hardware implementations given RNN model, including processing element design/optimization, quantization, activation implementation, etc. Experimental results on actual FPGA deployments show that E-RNN achieves a maximum energy efficiency improvement of 37.4$\times$ compared with ESE, and more than 2$\times$ compared with C-LSTM, under the same accuracy.