Recently, video streams have occupied a large proportion of Internet traffic, most of which contain human faces. Hence, it is necessary to predict saliency on multiple-face videos, which can provide attention cues for many content based applications. However, most of multiple-face saliency prediction works only consider visual information and ignore audio, which is not consistent with the naturalistic scenarios. Several behavioral studies have established that sound influences human attention, especially during the speech turn-taking in multiple-face videos. In this paper, we thoroughly investigate such influences by establishing a large-scale eye-tracking database of Multiple-face Video in Visual-Audio condition (MVVA). Inspired by the findings of our investigation, we propose a novel multi-modal video saliency model consisting of three branches: visual, audio and face. The visual branch takes the RGB frames as the input and encodes them into visual feature maps. The audio and face branches encode the audio signal and multiple cropped faces, respectively. A fusion module is introduced to integrate the information from three modalities, and to generate the final saliency map. Experimental results show that the proposed method outperforms 11 state-of-the-art saliency prediction works. It performs closer to human multi-modal attention.
Due to the rapid emergence of short videos and the requirement for content understanding and creation, the video captioning task has received increasing attention in recent years. In this paper, we convert traditional video captioning task into a new paradigm, \ie, Open-book Video Captioning, which generates natural language under the prompts of video-content-relevant sentences, not limited to the video itself. To address the open-book video captioning problem, we propose a novel Retrieve-Copy-Generate network, where a pluggable video-to-text retriever is constructed to retrieve sentences as hints from the training corpus effectively, and a copy-mechanism generator is introduced to extract expressions from multi-retrieved sentences dynamically. The two modules can be trained end-to-end or separately, which is flexible and extensible. Our framework coordinates the conventional retrieval-based methods with orthodox encoder-decoder methods, which can not only draw on the diverse expressions in the retrieved sentences but also generate natural and accurate content of the video. Extensive experiments on several benchmark datasets show that our proposed approach surpasses the state-of-the-art performance, indicating the effectiveness and promising of the proposed paradigm in the task of video captioning.
The Analog Ensemble (AnEn) technique has been shown effective on several weather problems. Unlike previous weather analogs that are sought within a large spatial domain and an extended temporal window, AnEn strictly confines space and time, and independently generates results at each grid point within a short time window. AnEn can find similar forecasts that lead to accurate and calibrated ensemble forecasts. The central core of the AnEn technique is a similarity metric that sorts historical forecasts with respect to a new target prediction. A commonly used metric is Euclidean distance. However, a significant difficulty using this metric is the definition of the weights for all the parameters. Generally, feature selection and extensive weight search are needed. This paper proposes a novel definition of weather analogs through a Machine Learning (ML) based similarity metric. The similarity metric uses neural networks that are trained and instantiated to search for weather analogs. This new metric allows incorporating all variables without requiring a prior feature selection and weight optimization. Experiments are presented on the application of this new metric to forecast wind speed and solar irradiance. Results show that the ML metric generally outperforms the original metric. The ML metric has a better capability to correct for larger errors and to take advantage of a larger search repository. Spatial predictions using a learned metric also show the ability to define effective latent features that are transferable to other locations.
The rising temperature is one of the key indicators of a warming climate, and it can cause extensive stress to biological systems as well as built structures. Due to the heat island effect, it is most severe in urban environments compared to other landscapes due to the decrease in vegetation associated with a dense human-built environment. It is essential to adequately monitor the local temperature dynamics to mitigate risks associated with increasing temperatures, which can include short term strategy to protect people and animals, to long term strategy to how to build a new structure and cope with extreme events. Observed temperature is also a very important input for atmospheric models, and accurate data can lead to better future forecasts. Ambient temperature collected at ground level can have a higher variability when compared to regional weather forecasts, which fail to capture the local dynamics. There remains a clear need for an accurate air temperature prediction at the sub-urban scale at high temporal and spatial resolution. This research proposed a framework based on Long Short-Term Memory (LSTM) deep learning network to generate day-ahead hourly temperature forecast with high spatial resolution. A case study is shown which uses historical in-situ observations and Internet of Things (IoT) observations for New York City, USA. By leveraging the historical air temperature data from in-situ observations, the LSTM model can be exposed to more historical patterns that might not be present in the IoT observations. Meanwhile, by using IoT observations, the spatial resolution of air temperature predictions is significantly improved.
The encoding of the target in object tracking moves from the coarse bounding-box to fine-grained segmentation map recently. Revisiting de facto real-time approaches that are capable of predicting mask during tracking, we observed that they usually fork a light branch from the backbone network for segmentation. Although efficient, directly fusing backbone features without considering the negative influence of background clutter tends to introduce false-negative predictions, lagging the segmentation accuracy. To mitigate this problem, we propose an attention retrieval network (ARN) to perform soft spatial constraints on backbone features. We first build a look-up-table (LUT) with the ground-truth mask in the starting frame, and then retrieves the LUT to obtain an attention map for spatial constraints. Moreover, we introduce a multi-resolution multi-stage segmentation network (MMS) to further weaken the influence of background clutter by reusing the predicted mask to filter backbone features. Our approach set a new state-of-the-art on recent pixel-wise object tracking benchmark VOT2020 while running at 40 fps. Notably, the proposed model surpasses SiamMask by 11.7/4.2/5.5 points on VOT2020, DAVIS2016, and DAVIS2017, respectively. We will release our code at https://github.com/researchmm/TracKit.
Anchor-based Siamese trackers have achieved remarkable advancements in accuracy, yet the further improvement is restricted by the lagged tracking robustness. We find the underlying reason is that the regression network in anchor-based methods is only trained on the positive anchor boxes (i.e., $IoU \geq0.6$). This mechanism makes it difficult to refine the anchors whose overlap with the target objects are small. In this paper, we propose a novel object-aware anchor-free network to address this issue. First, instead of refining the reference anchor boxes, we directly predict the position and scale of target objects in an anchor-free fashion. Since each pixel in groundtruth boxes is well trained, the tracker is capable of rectifying inexact predictions of target objects during inference. Second, we introduce a feature alignment module to learn an object-aware feature from predicted bounding boxes. The object-aware feature can further contribute to the classification of target objects and background. Moreover, we present a novel tracking framework based on the anchor-free model. The experiments show that our anchor-free tracker achieves state-of-the-art performance on five benchmarks, including VOT-2018, VOT-2019, OTB-100, GOT-10k and LaSOT. The source code is available at https://github.com/researchmm/TracKit.
Taking full advantage of the information from both vision and language is critical for the video captioning task. Existing models lack adequate visual representation due to the neglect of interaction between object, and sufficient training for content-related words due to long-tailed problems. In this paper, we propose a complete video captioning system including both a novel model and an effective training strategy. Specifically, we propose an object relational graph (ORG) based encoder, which captures more detailed interaction features to enrich visual representation. Meanwhile, we design a teacher-recommended learning (TRL) method to make full use of the successful external language model (ELM) to integrate the abundant linguistic knowledge into the caption model. The ELM generates more semantically similar word proposals which extend the ground-truth words used for training to deal with the long-tailed problem. Experimental evaluations on three benchmarks: MSVD, MSR-VTT and VATEX show the proposed ORG-TRL system achieves state-of-the-art performance. Extensive ablation studies and visualizations illustrate the effectiveness of our system.
Object detection and instance segmentation are two fundamental computer vision tasks. They are closely correlated but their relations have not yet been fully explored in most previous work. This paper presents RDSNet, a novel deep architecture for reciprocal object detection and instance segmentation. To reciprocate these two tasks, we design a two-stream structure to learn features on both the object level (i.e., bounding boxes) and the pixel level (i.e., instance masks) jointly. Within this structure, information from the two streams is fused alternately, namely information on the object level introduces the awareness of instance and translation variance to the pixel level, and information on the pixel level refines the localization accuracy of objects on the object level in return. Specifically, a correlation module and a cropping module are proposed to yield instance masks, as well as a mask based boundary refinement module for more accurate bounding boxes. Extensive experimental analyses and comparisons on the COCO dataset demonstrate the effectiveness and efficiency of RDSNet. The source code is available at https://github.com/wangsr126/RDSNet.