Italy

Both two-stream ConvNet and 3D ConvNet are widely used in action recognition. However, both methods are not efficient for deployment: calculating optical flow is very slow, while 3D convolution is computationally expensive. Our key insight is that the motion information from optical flow maps is complementary to the motion information from 3D ConvNet. Instead of simply combining these two methods, we propose two novel techniques to enhance the performance with less computational cost: fixed-motion-accumulation and balanced-motion-policy. With these two techniques, we propose a novel framework called Efficient Motion Complementary Network(EMC-Net) that enjoys both high efficiency and high performance. We conduct extensive experiments on Kinetics, UCF101, and Jester datasets. We achieve notably higher performance while consuming 4.7 times less computation than I3D, 11.6 times less computation than ECO, 17.8 times less computation than R(2+1)D. On Kinetics dataset, we achieve 2.6% better performance than the recent proposed TSM with 1.4 times fewer FLOPs and 10ms faster on K80 GPU.

ICPR 2020

Motion Complementary Network for Efficient Action Recognition

motion vector

effcient network

action recognition

ICPR 2020 is the premier international conference on Pattern Recognition, gathering leading researchers from around the world to present their latest achievements in Pattern Analysis and Recognition, applied Artificial Intelligence and Machine Learning. The topic of the year is " Putting Artificial intelligence to work on Patterns". 
For the first time, ICPR has employed a two-round review process similar to journal submissions, with new papers allowed to be submitted in either the first or second round and papers submitted in the first round and not accepted allowed to be revised and re-submitted for second round review. 

In this edition of the **ICPR 2020** Conference, there are 1407 Oral/Poster presentations organized in five Tracks: 
- Artificial Intelligence, Machine Learning for Pattern Analysis; 
- Biometrics, Human Analysis and Behavior Understanding; Computer Vision, Robotics and Intelligent Systems; 
- Document and Media Analysis; and 
- Image and Signal Processing. 

The **ICPR 2020** Main Conference program also includes 7 Keynotes from prominent world-renowned scientists, 8 Tutorials, 28 Demos, 8 Challenges, 2 Panels, as well as an Industry/Sponsor Exhibition. In addition, 41 Workshops on a range of important current and emerging research topics complete the ICPR 2020 program. 


poster


Hashing has been widely studied for cross-modal retrieval due to its promising efficiency and effectiveness in massive data analysis. However, most existing supervised hashing has the limitations of inefficiency for very large-scale search and intractable discrete constraint for hash codes learning. In this paper, we propose a new supervised hashing method, namely, Discrete Semantic Matrix Factorization Hashing (DSMFH), for cross-modal retrieval. First, we conduct the matrix factorization via directly utilizing the available label information to obtain a latent representation, so that both the inter-modality and intra-modality similarities are well preserved. Then, we simultaneously learn the discriminative hash codes and corresponding hash functions by deriving the matrix factorization into a discrete optimization. Finally, we adopt an alternatively iterative procedure to efficiently optimize the matrix factorization and discrete learning. Extensive experimental results on three widely used image-tag databases demonstrate the superiority of the DSMFH over state-of-the-art cross-modal hashing methods.

Discrete Semantic Matrix Factorization Hashing for Cross-Modal Retrieval

Subspace clustering is a useful technique for many computer vision applications in which the intrinsic dimension of high-dimensional data is often smaller than the ambient dimension. Spectral clustering, as one of the main approaches to subspace clustering, often takes on a sparse representation or a low-rank representation to learn a block diagonal self-representation matrix for subspace generation. However, existing methods require solving a large scale convex optimization problem with a large set of data, with computational complexity reaches O(N^3) for N data points. Therefore, the efficiency and scalability of traditional spectral clustering methods can not be guaranteed for large scale datasets. In this paper, we propose a subspace clustering model based on the Kronecker product. Due to the property that the Kronecker product of a block diagonal matrix with any other matrix is still a block diagonal matrix, we can efficiently learn the representation matrix which is formed by the Kronecker product of k smaller matrices. By doing so, our model significantly reduces the computational complexity to O(kN^(3/k)). Furthermore, our model is general in nature and can be adapted to different regularization based subspace clustering methods. Experimental results on two public datasets show that our model significantly improves the efficiency compared with several state-of-the-art methods. Moreover, we have conducted experiments on synthetic data to verify the scalability of our model for large scale datasets.

Fast Subspace Clustering Based on the Kronecker Product

Deep neural networks (DNN) have been applied recently to different domains and
perform better than classical state-of-the-art methods. However the high level of 
performances of DNNs is most often obtained with networks 
containing millions of parameters and for which training requires 
substantial computational power. 
To deal with this computational issue proximal regularization methods have been proposed in the literature but they are time consuming.\\
In this paper, we propose instead a constrained approach. 
We studied algorithms for different constraints: the classical $\ell_1$ unstructured constraint and structured constraints such as the $\ell_{2,1} $ constraint (Group LASSO). We propose a new $\ell_{1,1} $ structured constraint for which we provide a new projection algorithm. 
Finally, we used the recent "Lottery optimizer" replacing the threshold by our $\ell_{1,1} $ projection.
We demonstrate the effectiveness of this method with three popular datasets (MNIST, Fashion MNIST and CIFAR).
Experiments with these datasets show that our projection method using this new $\ell_{1,1} $ structured constraint provides the best decrease in memory and computational power.

Learning sparse deep neural networks using efficient structured projections on convex constraints for green AI

For the action localization task, pre-defined action anchors are the cornerstone of mainstream techniques. State-ofthe-art models mostly rely on a dense segmenting scheme, where anchors are sampled uniformly over the temporal domain with a predefined set of scales. However, it is not sufficient because action duration varies greatly. Therefore, it is necessary for the anchors or proposals to have a variable receptive field. In this paper, we propose a method called YOLA (You Ought to Look Around) which includes three parts: 1) a robust backbone SPN-I3D for extracting spatio-temporal features. In this part, we employ a stronger backbone I3D with SPN (Segment Pyramid Network) instead of C3D to obtain multi-scale features; 2) a simple but useful feature fusion module named LFE (Local Feature Extraction). Compared with the fully connected layer and global average pooling, our LFE model is more advantageous for network to fit and fuse features. 3) a new feature segment aligning method called TPGC (Two Pathway Graph Convolution), which allows one proposal to leverage semantic features of adjacent proposals to update its content and make sure the proposals have a variable receptive field. YOLA add only a small overhead to the baseline network, and is easy to train in an end-to-end manner, running at a speed of 1097 fps. YOLA achieves a mAP of 58.3%, outperforming all existing models including both RGBbased and two stream on THUMOS’14, and achieves competitive results on ActivityNet 1.3.

You Ought to Look Around: Precise, Large Span Action Detection


Clothing color transformation refers to changing the clothes color in an original image to the clothes color in a target image. In this paper, we propose an Unsupervised Clothing Color Transformation Generative Adversarial Network (UCCTGAN) for the task. UCCTGAN adopts the color histogram of a target clothes as color guidance and an improved U-net architecture called AntennaNet is put forward to fuse the extracted color information with the original image. Meanwhile, to accomplish unsupervised learning, the loss function is carefully designed according to color moment, which evaluates the chromatic aberration between the target clothing and the generated clothing. Experimental results show that our network has the ability to generate convincing color transformation results.

UCCTGAN: Unsupervised Clothing Color Transformation Generative Adversarial Network

Automatic cell segmentation is an essential step in the pipeline of computer-aided diagnosis (CAD), such as the detection and grading of breast cancer. Accurate segmentation of cells can not only assist the pathologists to make a more precise diagnosis, but also save much time and labor. However, this task suffers from stain variation, cell inhomogeneous intensities, background clutters and cells from different tissues. To address these issues, we propose an Attention Enforced Network (AENet), which is built on spatial attention module and channel attention module, to integrate local features with global dependencies and weight effective channels adaptively. Besides, we introduce a feature fusion branch to bridge high-level and low-level features. Finally, the marker controlled watershed algorithm is applied to post-process the predicted segmentation maps for reducing the fragmented regions. In the test stage, we present an individual color normalization method to deal with the stain variation problem. We evaluate this model on the MoNuSeg dataset. The quantitative comparisons against several prior methods demonstrate the superiority of our approach.

Accurate Cell Segmentation in Digital Pathology Images via Attention Enforced Networks


The architecture of the convolutional neural network (CNN) plays an important role in single image super-resolution (SISR). However, most models proposed in recent years usually transplant methods or architectures that perform well in other vision fields. Thence they do not combine the characteristics of super-resolution (SR) and ignore the key information brought by the recurring texture feature in the image. To utilize patch-recurrence in SR and the high correlation of texture, we propose a residual fractal convolutional block (RFCB) and expand its depth and width to obtain residual fractal network (RFN), which contains two variations, deep residual fractal network (DRFN) and wide residual fractal network (WRFN). RFCB is recursive with multiple branches of magnified receptive field. Through the phased feature fusion module, the network focuses on extracting high-frequency texture feature that repeatedly appear in the image. We also introduce residual in residual (RIR) structure to RFCB that enables abundant low-frequency feature feed into deeper layers and reduce the difficulties of network training. RFN is the first supervised learning method to combine the patch-recurrence characteristic in SISR into network design. Extensive experiments demonstrate that RFN outperforms state-of-the-art SISR methods in terms of both quantitative metrics and visual quality, while the amount of parameters has been greatly optimized. The source code and pre-trained models are released in https://github.com/JiahangGu/RFN.

Residual Fractal Network for Single Image Super Resolution by Widening and Deepening


The deep learning technique has recently led to significant improvement in object detection accuracy. In many applications, object detection is performed on video data consisting of a sequence of two-dimensional (2D) image frames. Numerous object detection schemes have been designed to detect objects independently in each video frame. Though temporal information within adjacent image frames can be exploited in subsequent object tracking stage, it has been shown that the object detection accuracy can be significantly improved by exploiting the temporal structure in the image sequence in the object detection stage. In this paper, we propose a novel video object detection method that exploits both the motion context inferred from the adjacent frames and the spatio-temporal features aggregated over the image sequence. First, correlation between the spatial feature maps over two adjacent frames are computed and the embedding vector, representing the motion context, is obtained by encoding the N correlation maps using long short term memory (LSTM). In addition to utilizing the motion context, the spatial feature maps for N+1 consecutive frames are aggregated to boost the quality of the feature map. The gated attention network is employed to selectively combine the temporal feature maps based on their relevance to the feature map in the present image frame. While most video object detectors have been developed for two-stage object detectors, our proposed idea applies to one-stage detectors with the advantage of low computational complexity in practical real-time applications. Our numerical evaluation conducted on the ImageNet object detection from video (VID) dataset demonstrates that our proposed network achieves significant performance gain over the baseline algorithms and outperforms the existing one-stage video object detectors.

Video Object Detection Using Object’s Motion Context and Spatio-Temporal Feature Aggregation


Learning-based stereo matching has recently achieved promising results, yet still suffers difficulties in establishing reliable matches in weakly matchable regions that are textureless, non-Lambertian, or occluded. In this paper, we address this challenge by proposing a stereo matching network that considers pixel-wise matchability. Specifically, the network jointly regresses disparity and matchability maps from 3D probability volume through expectation and entropy operations. Next, a learned attenuation is applied as the robust loss function to alleviate the influence of weakly matchable pixels in the training. Finally, a matchability-aware disparity refinement is introduced to improve the depth inference in weakly matchable regions. The proposed deep stereo matchability (DSM) framework can
improve the matching result or accelerate the computation while still guaranteeing the quality. Moreover, the DSM framework is portable to many recent stereo networks. Extensive experiments are conducted on Scene Flow and KITTI stereo datasets to demonstrate the effectiveness of the proposed framework over the state-of-the-art learning-based stereo methods.

Learning Stereo Matchability in Disparity Regression Networks

Facial actions are spatio-temporal signals by nature, and therefore their modeling is crucially dependent on the availability of temporal information. In this paper, we focus on inferring such temporal dynamics of facial actions when no explicit temporal information is available, i.e. from still images. We present a novel self-supervised learning approach to capture multiple scales of temporal dynamics, with an application to facial Action Unit (AU) intensity estimation and dimensional affect estimation. In particular: 1. We propose a framework that infers a dynamic representation (DR) from a still image, capturing the bi-directional flow of time within a short time-window centered at the input image; 2. We show that the proposed rank loss can apply facial temporal evolution to self-supervise the training process without using target representations, allowing the network to represent dynamics more broadly; 3. We propose a multiple temporal scale approach that infers DRs for different window lengths (MDR) from a still image. We empirically validate the value of our approach on the task of frame ranking, and show how our proposed MDR attains state of the art results on BP4D for AU intensity estimation and on SEMAINE for dimensional affect estimation, using only still images at test time.

Downloads

Next from ICPR 2020

Discrete Semantic Matrix Factorization Hashing for Cross-Modal Retrieval

Similar lecture

DeepPEAR: Deep Pose Estimation and Action Recognition