Video-based action recognition with deep neural networks has shown remarkable progress. However, most of the existing approaches are too computationally expensive due to the complex network architecture. To address these problems, we propose a new real-time action recognition architecture, called Temporal Convolutional 3D Network (T-C3D), which learns video action representations in a hierarchical multi-granularity manner. Specifically, we combine a residual 3D convolutional neural network which captures complementary information on the appearance of a single frame and the motion between consecutive frames with a new temporal encoding method to explore the temporal dynamics of the whole video. Thus heavy calculations are avoided when doing the inference, which enables the method to be capable of real-time processing. On two challenging benchmark datasets, UCF101 and HMDB51, our method is significantly better than state-of-the-art real-time methods by over 5.4% in terms of accuracy and 2 times faster in terms of inference speed (969 frames per second), demonstrating comparable recognition performance to the state-of-the-art methods. The source code for the complete system as well as the pre-trained models are publicly available at https://github.com/tc3d.

Action recognition is an important yet challenging task in computer vision. In this paper, we propose a novel deep-based framework for action recognition, which improves the recognition accuracy by: 1) deriving more precise features for representing actions, and 2) reducing the asynchrony between different information streams. We first introduce a coarse-to-fine network which extracts shared deep features at different action class granularities and progressively integrates them to obtain a more accurate feature representation for input actions. We further introduce an asynchronous fusion network. It fuses information from different streams by asynchronously integrating stream-wise features at different time points, hence better leveraging the complementary information in different streams. Experimental results on action recognition benchmarks demonstrate that our approach achieves the state-of-the-art performance.

Restoring face images from distortions is important in face recognition applications and is challenged by multiple scale issues, which is still not well-solved in research area. In this paper, we present a Sequential Gating Ensemble Network (SGEN) for multi-scale face restoration issue. We first employ the principle of ensemble learning into SGEN architecture design to reinforce predictive performance of the network. The SGEN aggregates multi-level base-encoders and base-decoders into the network, which enables the network to contain multiple scales of receptive field. Instead of combining these base-en/decoders directly with non-sequential operations, the SGEN takes base-en/decoders from different levels as sequential data. Specifically, the SGEN learns to sequentially extract high level information from base-encoders in bottom-up manner and restore low level information from base-decoders in top-down manner. Besides, we propose to realize bottom-up and top-down information combination and selection with Sequential Gating Unit (SGU). The SGU sequentially takes two inputs from different levels and decides the output based on one active input. Experiment results demonstrate that our SGEN is more effective at multi-scale human face restoration with more image details and less noise than state-of-the-art image restoration models. By using adversarial training, SGEN also produces more visually preferred results than other models through subjective evaluation.

Cross-view person identification (CVPI) from multiple temporally synchronized videos taken by multiple wearable cameras from different, varying views is a very challenging but important problem, which has attracted more interests recently. Current state-of-the-art performance of CVPI is achieved by matching appearance and motion features across videos, while the matching of pose features does not work effectively given the high inaccuracy of the 3D human pose estimation on videos/images collected in the wild. In this paper, we introduce a new metric of confidence to the 3D human pose estimation and show that the combination of the inaccurately estimated human pose and the inferred confidence metric can be used to boost the CVPI performance---the estimated pose information can be integrated to the appearance and motion features to achieve the new state-of-the-art CVPI performance. More specifically, the estimated confidence metric is measured at each human-body joint and the joints with higher confidence are weighted more in the pose matching for CVPI. In the experiments, we validate the proposed method on three wearable-camera video datasets and compare the performance against several other existing CVPI methods.

The recent development of CNN-based image dehazing has revealed the effectiveness of end-to-end modeling. However, extending the idea to end-to-end video dehazing has not been explored yet. In this paper, we propose an End-to-End Video Dehazing Network (EVD-Net), to exploit the temporal consistency between consecutive video frames. A thorough study has been conducted over a number of structure options, to identify the best temporal fusion strategy. Furthermore, we build an End-to-End United Video Dehazing and Detection Network (EVDD-Net), which concatenates and jointly trains EVD-Net with a video object detection model. The resulting augmented end-to-end pipeline has demonstrated much more stable and accurate detection results in hazy video.

Tracking target of interests is an important step for motion perception in intelligent video surveillance systems. While most recently developed tracking algorithms are grounded in RGB image sequences, it should be noted that information from RGB modality is not always reliable (e.g. in a dark environment with poor lighting condition), which urges the need to integrate information from infrared modality for effective tracking because of the insensitivity to illumination condition of infrared thermal camera. However, several issues encountered during the tracking process limit the fusing performance of these heterogeneous modalities: 1) the cross-modality discrepancy of visual and motion characteristics, 2) the uncertainty of degree of reliability in different modalities, and 3) large target appearance variations and background distractions within each modality. To address these issues, this paper proposes a novel and optimal discriminative learning framework for multi-modality tracking. In particular, the proposed discriminative learning framework is able to: 1) jointly eliminate outlier samples caused by large variations and learn discriminability-consistent features from heterogeneous modalities, and 2) collaboratively perform modality reliability measurement and target-background separation. Extensive experiments on RGB-infrared image sequences demonstrate the effectiveness of the proposed method.

Existing action detection algorithms usually generate action proposals through an extensive search over the video at multiple temporal scales, which brings about huge computational overhead and deviates from the human perception procedure. We argue that the process of detecting actions should be naturally one of observation and refinement: observe the current window and refine the span of attended window to cover true action regions. In this paper, we propose a Self-Adaptive Proposal (SAP) model that learns to find actions through continuously adjusting the temporal bounds in a self-adaptive way. The whole process can be deemed as an agent, which is firstly placed at the beginning of the video and traverse the whole video by adopting a sequence of transformations on the current attended region to discover actions according to a learned policy. We utilize reinforcement learning, especially the Deep Q-learning algorithm to learn the agent’s decision policy. In addition, we use temporal pooling operation to extract more effective feature representation for the long temporal window, and design a regression network to adjust the position offsets between predicted results and the ground truth. Experiment results on THUMOS’14 validate the effectiveness of SAP, which can achieve competitive performance with current action detection algorithms via much fewer proposals.

Salient object detection is a fundamental yet challenging problem in computer vision, aiming to highlight the most visually distinctive objects or regions in an image. Recent works benefit from the development of fully convolutional neural networks (FCNs) and achieve great success by integrating features from multiple layers of FCNs. However, the integrated features tend to include non-salient regions (due to low level features of the FCN) or lost details of salient objects (due to high level features of the FCN) when producing the saliency maps. In this paper, we develop a novel deep saliency network equipped with recurrently aggregated deep features (RADF) to more accurately detect salient objects from an image by fully exploiting the complementary saliency information captured in different layers. The RADF utilizes the multi-level features integrated from different layers of a FCN to recurrently refine the features at each layer, suppressing the non-salient noise at low-level of the FCN and increasing more salient details into features at high layers. We perform experiments to evaluate the effectiveness of the proposed network on 5 famous saliency detection benchmarks and compare it with 15 state-of-the-art methods. Our method ranks first in 4 of the 5 datasets and second in the left dataset.

Cascaded Regression (CR) based methods have been proposed to solve facial landmarks detection problem, which learn a series of descent directions by multiple cascaded regressors separately trained in coarse and fine stages. They outperform the traditional gradient descent based methods in both accuracy and running speed. However, cascaded regression is not robust enough because each regressor's training data comes from the output of previous regressor. Moreover, training multiple regressors requires lots of computing resources, especially for deep learning based methods. In this paper, we develop a Self-Iterative Regression (SIR) framework to improve the model efficiency. Only one self-iterative regressor is trained to learn the descent directions for samples from coarse stages to fine stages, and parameters are iteratively updated by the same regressor. Specifically, we proposed Landmarks-Attention Network (LAN) as our regressor, which concurrently learns features around each landmark and obtains the holistic location increment. By doing so, not only the rest of regressors are removed to simplify the training process, but the number of model parameters is significantly decreased. The experiments demonstrate that with only 3.72M model parameters, our proposed method achieves the state-of-the-art performance.

Zero-shot learning (ZSL) is regarded as an effective way to construct classification models for target classes which have no labeled samples available. The basic framework is to transfer knowledge from (different) auxiliary source classes having sufficient labeled samples with some attributes shared by target and source classes as bridge. Attributes play an important role in ZSL but they have not gained sufficient attention in recent years. Previous works mostly assume attributes are perfect and treat each attribute equally. However, as shown in this paper, different attributes have different properties, such as their class distribution, variance, and entropy, which may have considerable impact on ZSL accuracy if treated equally. Based on this observation, in this paper we propose to use a subset of attributes, instead of the whole set, for building ZSL models. The attribute selection is conducted by considering the information amount and predictability under a novel joint optimization framework. To our knowledge, this is the first work that notices the influence of attributes themselves and proposes to use a refined attribute set for ZSL. Since our approach focuses on selecting good attributes for ZSL, it can be combined to any attribute based ZSL approaches so as to augment their performance. Experiments on four ZSL benchmarks demonstrate that our approach can improve zero-shot classification accuracy and yield state-of-the-art results.