Recently, video classification attracts intensive research efforts. However, most existing works are based on framelevel visual features, which might fail to model the temporal information, e.g. characteristics accumulated along time. In order to capture video temporal information, we propose to analyse features in frequency domain transformed by discrete Fourier transform (DFT features). Frame-level features are firstly extract by a pre-trained deep convolutional neural network (CNN). Then, time domain features are transformed and interpolated into DFT features. CNN and DFT features are further encoded by using different pooling methods and fused for video classification. In this way, static image features extracted from a pre-trained deep CNN and temporal information represented by DFT features are jointly considered for video classification. We test our method for video emotion classification and action recognition. Experimental results demonstrate that combining DFT features can effectively capture temporal information and therefore improve the performance of both video emotion classification and action recognition. Our approach has achieved a state-of-the-art performance on the largest video emotion dataset (VideoEmotion-8 dataset) and competitive results on UCF-101.
Action recognition in videos is a challenging task due to the complexity of the spatio-temporal patterns to model and the difficulty to acquire and learn on large quantities of video data. Deep learning, although a breakthrough for image classification and showing promise for videos, has still not clearly superseded action recognition methods using hand-crafted features, even when training on massive datasets. In this paper, we introduce hybrid video classification architectures based on carefully designed unsupervised representations of hand-crafted spatio-temporal features classified by supervised deep networks. As we show in our experiments on five popular benchmarks for action recognition, our hybrid model combines the best of both worlds: it is data efficient (trained on 150 to 10000 short clips) and yet improves significantly on the state of the art, including recent deep models trained on millions of manually labelled images and videos.
The video and action classification have extremely evolved by deep neural networks specially with two stream CNN using RGB and optical flow as inputs and they present outstanding performance in terms of video analysis. One of the shortcoming of these methods is handling motion information extraction which is done out side of the CNNs and relatively time consuming also on GPUs. So proposing end-to-end methods which are exploring to learn motion representation, like 3D-CNN can achieve faster and accurate performance. We present some novel deep CNNs using 3D architecture to model actions and motion representation in an efficient way to be accurate and also as fast as real-time. Our new networks learn distinctive models to combine deep motion features into appearance model via learning optical flow features inside the network.
Many recent advancements in Computer Vision are attributed to large datasets. Open-source software packages for Machine Learning and inexpensive commodity hardware have reduced the barrier of entry for exploring novel approaches at scale. It is possible to train models over millions of examples within a few days. Although large-scale datasets exist for image understanding, such as ImageNet, there are no comparable size video classification datasets. In this paper, we introduce YouTube-8M, the largest multi-label video classification dataset, composed of ~8 million videos (500K hours of video), annotated with a vocabulary of 4800 visual entities. To get the videos and their labels, we used a YouTube video annotation system, which labels videos with their main topics. While the labels are machine-generated, they have high-precision and are derived from a variety of human-based signals including metadata and query click signals. We filtered the video labels (Knowledge Graph entities) using both automated and manual curation strategies, including asking human raters if the labels are visually recognizable. Then, we decoded each video at one-frame-per-second, and used a Deep CNN pre-trained on ImageNet to extract the hidden representation immediately prior to the classification layer. Finally, we compressed the frame features and make both the features and video-level labels available for download. We trained various (modest) classification models on the dataset, evaluated them using popular evaluation metrics, and report them as baselines. Despite the size of the dataset, some of our models train to convergence in less than a day on a single machine using TensorFlow. We plan to release code for training a TensorFlow model and for computing metrics.
The CNN-encoding of features from entire videos for the representation of human actions has rarely been addressed. Instead, CNN work has focused on approaches to fuse spatial and temporal networks, but these were typically limited to processing shorter sequences. We present a new video representation, called temporal linear encoding (TLE) and embedded inside of CNNs as a new layer, which captures the appearance and motion throughout entire videos. It encodes this aggregated information into a robust video feature representation, via end-to-end learning. Advantages of TLEs are: (a) they encode the entire video into a compact feature representation, learning the semantics and a discriminative feature space; (b) they are applicable to all kinds of networks like 2D and 3D CNNs for video classification; and (c) they model feature interactions in a more expressive way and without loss of information. We conduct experiments on two challenging human action datasets: HMDB51 and UCF101. The experiments show that TLE outperforms current state-of-the-art methods on both datasets.
Robust visual tracking is a challenging computer vision problem, with many real-world applications. Most existing approaches employ hand-crafted appearance features, such as HOG or Color Names. Recently, deep RGB features extracted from convolutional neural networks have been successfully applied for tracking. Despite their success, these features only capture appearance information. On the other hand, motion cues provide discriminative and complementary information that can improve tracking performance. Contrary to visual tracking, deep motion features have been successfully applied for action recognition and video classification tasks. Typically, the motion features are learned by training a CNN on optical flow images extracted from large amounts of labeled videos. This paper presents an investigation of the impact of deep motion features in a tracking-by-detection framework. We further show that hand-crafted, deep RGB, and deep motion features contain complementary information. To the best of our knowledge, we are the first to propose fusing appearance information with deep motion features for visual tracking. Comprehensive experiments clearly suggest that our fusion approach with deep motion features outperforms standard methods relying on appearance information alone.
Encoding time-series with Linear Dynamical Systems (LDSs) leads to rich models with applications ranging from dynamical texture recognition to video segmentation to name a few. In this paper, we propose to represent LDSs with infinite-dimensional subspaces and derive an analytic solution to obtain stable LDSs. We then devise efficient algorithms to perform sparse coding and dictionary learning on the space of infinite-dimensional subspaces. In particular, two solutions are developed to sparsely encode an LDS. In the first method, we map the subspaces into a Reproducing Kernel Hilbert Space (RKHS) and achieve our goal through kernel sparse coding. As for the second solution, we propose to embed the infinite-dimensional subspaces into the space of symmetric matrices and formulate the sparse coding accordingly in the induced space. For dictionary learning, we encode time-series by introducing a novel concept, namely the two-fold LDSs. We then make use of the two-fold LDSs to derive an analytical form for updating atoms of an LDS dictionary, i.e., each atom is an LDS itself. Compared to several baselines and state-of-the-art methods, the proposed methods yield higher accuracies in various classification tasks including video classification and tactile recognition.
Learning transformation invariant representations of visual data is an important problem in computer vision. Deep convolutional networks have demonstrated remarkable results for image and video classification tasks. However, they have achieved only limited success in the classification of images that undergo geometric transformations. In this work we present a novel Transformation Invariant Graph-based Network (TIGraNet), which learns graph-based features that are inherently invariant to isometric transformations such as rotation and translation of input images. In particular, images are represented as signals on graphs, which permits to replace classical convolution and pooling layers in deep networks with graph spectral convolution and dynamic graph pooling layers that together contribute to invariance to isometric transformations. Our experiments show high performance on rotated and translated images from the test set compared to classical architectures that are very sensitive to transformations in the data. The inherent invariance properties of our framework provide key advantages, such as increased resiliency to data variability and sustained performance with limited training sets.
Video classification is productive in many practical applications, and the recent deep learning has greatly improved its accuracy. However, existing works often model video frames indiscriminately, but from the view of motion, video frames can be decomposed into salient and non-salient areas naturally. Salient and non-salient areas should be modeled with different networks, for the former present both appearance and motion information, and the latter present static background information. To address this problem, in this paper, video saliency is predicted by optical flow without supervision firstly. Then two streams of 3D CNN are trained individually for raw frames and optical flow on salient areas, and another 2D CNN is trained for raw frames on non-salient areas. For the reason that these three streams play different roles for each class, the weights of each stream are adaptively learned for each class. Experimental results show that saliency-guided modeling and adaptively weighted learning can reinforce each other, and we achieve the state-of-the-art results.
Recent two-stream deep Convolutional Neural Networks (ConvNets) have made significant progress in recognizing human actions in videos. Despite their success, methods extending the basic two-stream ConvNet have not systematically explored possible network architectures to further exploit spatiotemporal dynamics within video sequences. Further, such networks often use different baseline two-stream networks. Therefore, the differences and the distinguishing factors between various methods using Recurrent Neural Networks (RNN) or convolutional networks on temporally-constructed feature vectors (Temporal-ConvNet) are unclear. In this work, we first demonstrate a strong baseline two-stream ConvNet using ResNet-101. We use this baseline to thoroughly examine the use of both RNNs and Temporal-ConvNets for extracting spatiotemporal information. Building upon our experimental results, we then propose and investigate two different networks to further integrate spatiotemporal information: 1) temporal segment RNN and 2) Inception-style Temporal-ConvNet. We demonstrate that using both RNNs (using LSTMs) and Temporal-ConvNets on spatiotemporal feature matrices are able to exploit spatiotemporal dynamics to improve the overall performance. However, each of these methods require proper care to achieve state-of-the-art performance; for example, LSTMs require pre-segmented data or else they cannot fully exploit temporal information. Our analysis identifies specific limitations for each method that could form the basis of future work. Our experimental results on UCF101 and HMDB51 datasets achieve state-of-the-art performances, 94.1% and 69.0%, respectively, without requiring extensive temporal augmentation.
In this work, we introduce a new video representation for action classification that aggregates local convolutional features across the entire spatio-temporal extent of the video. We do so by integrating state-of-the-art two-stream networks with learnable spatio-temporal feature aggregation. The resulting architecture is end-to-end trainable for whole-video classification. We investigate different strategies for pooling across space and time and combining signals from the different streams. We find that: (i) it is important to pool jointly across space and time, but (ii) appearance and motion streams are best aggregated into their own separate representations. Finally, we show that our representation outperforms the two-stream base architecture by a large margin (13% relative) as well as out-performs other baselines with comparable base architectures on HMDB51, UCF101, and Charades video classification benchmarks.
Deep convolutional networks have achieved great success for image recognition. However, for action recognition in videos, their advantage over traditional methods is not so evident. We present a general and flexible video-level framework for learning action models in videos. This method, called temporal segment network (TSN), aims to model long-range temporal structures with a new segment-based sampling and aggregation module. This unique design enables our TSN to efficiently learn action models by using the whole action videos. The learned models could be easily adapted for action recognition in both trimmed and untrimmed videos with simple average pooling and multi-scale temporal window integration, respectively. We also study a series of good practices for the instantiation of TSN framework given limited training samples. Our approach obtains the state-the-of-art performance on four challenging action recognition benchmarks: HMDB51 (71.0%), UCF101 (94.9%), THUMOS14 (80.1%), and ActivityNet v1.2 (89.6%). Using the proposed RGB difference for motion models, our method can still achieve competitive accuracy on UCF101 (91.0%) while running at 340 FPS. Furthermore, based on the temporal segment networks, we won the video classification track at the ActivityNet challenge 2016 among 24 teams, which demonstrates the effectiveness of TSN and the proposed good practices.
This article describes the final solution of team monkeytyping, who finished in second place in the YouTube-8M video understanding challenge. The dataset used in this challenge is a large-scale benchmark for multi-label video classification. We extend the work in [1] and propose several improvements for frame sequence modeling. We propose a network structure called Chaining that can better capture the interactions between labels. Also, we report our approaches in dealing with multi-scale information and attention pooling. In addition, We find that using the output of model ensemble as a side target in training can boost single model performance. We report our experiments in bagging, boosting, cascade, and stacking, and propose a stacking algorithm called attention weighted stacking. Our final submission is an ensemble that consists of 74 sub models, all of which are listed in the appendix.
Videos are inherently multimodal. This paper studies the problem of how to fully exploit the abundant multimodal clues for improved video categorization. We introduce a hybrid deep learning framework that integrates useful clues from multiple modalities, including static spatial appearance information, motion patterns within a short time window, audio information as well as long-range temporal dynamics. More specifically, we utilize three Convolutional Neural Networks (CNNs) operating on appearance, motion and audio signals to extract their corresponding features. We then employ a feature fusion network to derive a unified representation with an aim to capture the relationships among features. Furthermore, to exploit the long-range temporal dynamics in videos, we apply two Long Short Term Memory networks with extracted appearance and motion features as inputs. Finally, we also propose to refine the prediction scores by leveraging contextual relationships among video semantics. The hybrid deep learning framework is able to exploit a comprehensive set of multimodal features for video classification. Through an extensive set of experiments, we demonstrate that (1) LSTM networks which model sequences in an explicitly recurrent manner are highly complementary with CNN models; (2) the feature fusion network which produces a fused representation through modeling feature relationships outperforms alternative fusion strategies; (3) the semantic context of video classes can help further refine the predictions for improved performance. Experimental results on two challenging benchmarks, the UCF-101 and the Columbia Consumer Videos (CCV), provide strong quantitative evidence that our framework achieves promising results: $93.1\%$ on the UCF-101 and $84.5\%$ on the CCV, outperforming competing methods with clear margins.
Video classification problem has been studied many years. The success of Convolutional Neural Networks (CNN) in image recognition tasks gives a powerful incentive for researchers to create more advanced video classification approaches. As video has a temporal content Long Short Term Memory (LSTM) networks become handy tool allowing to model long-term temporal clues. Both approaches need a large dataset of input data. In this paper three models provided to address video classification using recently announced YouTube-8M large-scale dataset. The first model is based on frame pooling approach. Two other models based on LSTM networks. Mixture of Experts intermediate layer is used in third model allowing to increase model capacity without dramatically increasing computations. The set of experiments for handling imbalanced training data has been conducted.
The emergence of depth imaging technologies like the Microsoft Kinect has renewed interest in computational methods for gesture classification based on videos. For several years now, researchers have used the Bag-of-Features (BoF) as a primary method for generation of feature vectors from video data for recognition of gestures. However, the BoF method is a coarse representation of the information in a video, which often leads to poor similarity measures between videos. Besides, when features extracted from different spatio-temporal locations in the video are pooled to create histogram vectors in the BoF method, there is an intrinsic loss of their original locations in space and time. In this paper, we propose a new Multiresolution Match Kernel (MMK) for video classification, which can be considered as a generalization of the BoF method. We apply this procedure to hand gesture classification based on RGB-D videos of the American Sign Language(ASL) hand gestures and our results show promise and usefulness of this new method.
Large-scale datasets have played a significant role in progress of neural network and deep learning areas. YouTube-8M is such a benchmark dataset for general multi-label video classification. It was created from over 7 million YouTube videos (450,000 hours of video) and includes video labels from a vocabulary of 4716 classes (3.4 labels/video on average). It also comes with pre-extracted audio & visual features from every second of video (3.2 billion feature vectors in total). Google cloud recently released the datasets and organized 'Google Cloud & YouTube-8M Video Understanding Challenge' on Kaggle. Competitors are challenged to develop classification algorithms that assign video-level labels using the new and improved Youtube-8M V2 dataset. Inspired by the competition, we started exploration of audio understanding and classification using deep learning algorithms and ensemble methods. We built several baseline predictions according to the benchmark paper and public github tensorflow code. Furthermore, we improved global prediction accuracy (GAP) from base level 77% to 80.7% through approaches of ensemble.
Recent advances have enabled "oracle" classifiers that can classify across many classes and input distributions with high accuracy without retraining. However, these classifiers are relatively heavyweight, so that applying them to classify video is costly. We show that day-to-day video exhibits highly skewed class distributions over the short term, and that these distributions can be classified by much simpler models. We formulate the problem of detecting the short-term skews online and exploiting models based on it as a new sequential decision making problem dubbed the Online Bandit Problem, and present a new algorithm to solve it. When applied to recognizing faces in TV shows and movies, we realize end-to-end classification speedups of 2.4-7.8x/2.6-11.2x (on GPU/CPU) relative to a state-of-the-art convolutional neural network, at competitive accuracy.
This paper introduces the system we developed for the Google Cloud & YouTube-8M Video Understanding Challenge, which can be considered as a multi-label classification problem defined on top of the large scale YouTube-8M Dataset. We employ a large set of techniques to aggregate the provided frame-level feature representations and generate video-level predictions, including several variants of recurrent neural networks (RNN) and generalized VLAD. We also adopt several fusion strategies to explore the complementarity among the models. In terms of the official metric GAP@20 (global average precision at 20), our best fusion model attains 0.84198 on the public 50\% of test data and 0.84193 on the private 50\% of test data, ranking 4th out of 650 teams worldwide in the competition.
The Recurrent Neural Networks and their variants have shown promising performances in sequence modeling tasks such as Natural Language Processing. These models, however, turn out to be impractical and difficult to train when exposed to very high-dimensional inputs due to the large input-to-hidden weight matrix. This may have prevented RNNs' large-scale application in tasks that involve very high input dimensions such as video modeling; current approaches reduce the input dimensions using various feature extractors. To address this challenge, we propose a new, more general and efficient approach by factorizing the input-to-hidden weight matrix using Tensor-Train decomposition which is trained simultaneously with the weights themselves. We test our model on classification tasks using multiple real-world video datasets and achieve competitive performances with state-of-the-art models, even though our model architecture is orders of magnitude less complex. We believe that the proposed approach provides a novel and fundamental building block for modeling high-dimensional sequential data with RNN architectures and opens up many possibilities to transfer the expressive and advanced architectures from other domains such as NLP to modeling high-dimensional sequential data.