Italy

As a typical cross-modal problem, visual question answering (VQA) has received increasing attention from the communities of computer vision and natural language processing. Reading and reasoning about texts in the images is a burgeoning and important topic in VQA, especially for the visually impaired assistance applications. Given a image, it aims to predict an answer to a provided natural language question closely related to its textual contents. In this paper, we propose a novel end-to-end textual content based VQA model, which grounds question answering both on the visual and textual information. 


ICPR 2020

Dual Path Multi-modal High-Order Features for Textual Content based Visual Question Answering

semantic embeddings

multi-modal high-order features

textvqa

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. 


technical paper


Cancer is one of the leading causes of death in the western world, with medical imaging playing a key role for early diagnosis. Focusing on breast cancer, one of the emerging imaging methodologies is Dynamic Contrast Enhanced-Magnetic
Resonance Imaging (DCE-MRI). The flip side of using DCE-MRI is in its long acquisition times that often causes the patient to move. This results in motion artefacts, namely distortions in the acquired image that can affect DCE-MRI analysis. A possible solution consists in the use of Motion Correction Techniques
(MCTs), i.e. procedures intended to re-align the postcontrast image to the corresponding pre-contrast (reference) one. This task is particularly critic in DCE-MRI, due to brightness variations introduced in post-contrast images by the contrast agent flowing. To face this problem, in this work we introduce a new MCT for breast DCE-MRI leveraging Physiologically Based PharmacoKinetic (PBPK) modelling and Artificial Neural Networks (ANN) to determine the most suitable physiologically compliant transformation. To this aim, we propose a Neural Registration Network relying on a very task-specific loss function explicitly designed to take into account the contrast agent flowing while enforcing a correct re-alignment. We compared the obtained results against some conventional motion correction techniques, evaluating the performance on a patient-by-patient basis. Results show that the proposed approach results to be the best performing even when compared against other techniques designed to take into account for brightness variations.

Neural Machine Registration for Motion Correction in Breast DCE-MRI

Continuous hand gesture recognition (HGR) is an essential part of human-computer interaction with a wide range of applications in the automotive sector, consumer electronics, home automation, and others. In recent years, accurate and efficient deep learning models have been proposed for HGR. However, in the research community, the current publicly available datasets lack real-world elements needed to build responsive and efficient HGR systems. In this paper, we introduce a new benchmark dataset named IPN Hand with sufficient size, variety, and real-world elements able to train and evaluate deep neural networks. This dataset contains more than 4,000 gesture samples and 800,000 RGB frames from 50 distinct subjects. We design 13 different static and dynamic gestures focused on interaction with touchless screens. We especially consider the scenario when continuous gestures are performed without transition states, and when subjects perform natural movements with their hands as non-gesture actions. Gestures were collected from about 30 diverse scenes, with real-world variation in background and illumination. With our dataset, the performance of three 3D-CNN models is evaluated on the tasks of isolated and continuous real-time HGR. Furthermore, we analyze the possibility of increasing the recognition accuracy by adding multiple modalities derived from RGB frames, i.e., optical flow and semantic segmentation, while keeping the real-time performance of the 3D-CNN model. Our empirical study also provides a comparison with the publicly available nvGesture (NVIDIA) dataset. The experimental results show that the state-of-the-art ResNext-101 model decreases about 30% accuracy when using our real-world dataset, demonstrating that the IPN Hand dataset can be used as a benchmark, and may help the community to step forward in the continuous HGR.

IPN Hand: A Video Dataset and Benchmark for Real-Time Continuous Hand Gesture Recognition

We propose a rank-based metric learning method by leveraging a concept of the Riemannian Potato for better separating non-linear data. By exploring the geometric properties of Riemannian manifolds, the proposed loss function optimizes the measure of dispersion using the distribution of Riemannian distances between a reference sample and neighbors and builds a ranked list according to the similarities. We show the proposed function can learn a hypersphere for each class, preserving the similarity structure inside it on Riemannian manifold. As a result, compared with Euclidean distance-based metric, our method can further jointly reduce the intra-class distances and enlarge the inter-class distances for learned features, consistently outperforming state-of-the-art methods on three widely used non- linear datasets.

Nonlinear Ranking Loss on Riemannian Potato Embedding


Computer vision with state-of-the-art deep learning models has achieved huge success in the field of Optical Character Recognition (OCR) including text detection and recognition tasks recently. However, Key Information Extraction (KIE) from documents as the downstream task of OCR, having a large number of use scenarios in the real-world, remains a challenge because documents not only have textual features extracting from OCR systems but also have semantic visual features that are not fully exploited and play a critical role in KIE. Too little work has been devoted to efficiently make full use of both textual and visual features of the documents. In this paper, we introduce PICK, a framework that is effective and robust in handling complex document layout for KIE by combining graph learning with graph convolution operation, yielding a richer semantic representation containing the textual and visual features and global layout without ambiguity. Extensive experiments on real-world datasets have been conducted to show that our method outperforms baseline methods by significant margins. Our code is available at https://github.com/wenwenyu/PICK-pytorch.

PICK: Processing Key Information Extraction from Documents using Improved Graph Learning-Convolutional Networks

Video saliency detection is an important low-level task that has been used in a large range of high-level applications. In this paper, we proposed a three-stream multi-scale attentive network (TSMSAN) for saliency detection in dynamic scenes. TSMSAN integrates motion vector (MV) representation, static saliency map, and RGB information in multi-scales together into one framework on the basis of Fully Convolutional Network (FCN) and spatial attention mechanism. On the one hand, the respective motion features, spatial features, as well as scene features can provide abundant information for video saliency detection. On the other hand, spatial attention mechanism can combine features with multi-scales to focus on key information in dynamic scenes. In this manner, the proposed TSMSAN can encode the spatiotemporal features of the dynamic scene comprehensively.
We evaluate the proposed approach on two public dynamic saliency datasets. The experimental results demonstrate TSMSAN is able to achieve the state-of-the-art performance as well as excellent generalization ability. Furthermore, the proposed TSMSAN can provide more convincing video saliency information, in line with human perception.

TSMSAN: A Three-Stream Multi-Scale Attentive Network for Video Saliency Detection

Recent machine learning methods use increasingly large deep neural networks to achieve state-of-the-art results in various tasks. Network quantization can effectively reduce computation and memory costs without modifying network
structures, facilitating the deployment of deep neural networks (DNNs) on cloud and edge devices. However, most of the existing methods usually need time-consuming training or fine-tuning and access to the original training dataset that may be unavailable due to privacy or security concerns. In this paper, we present a novel
method to achieve low-precision quantization with limited data. Firstly, to reduce the complexity of per-channel quantization and degeneration of per-layer quantization, we introduce group quantization that separates the output channels into groups and processes each group independently. Secondly, to better distill
knowledge from the pre-trained FP32 model with limited data, we introduce a two-stage knowledge distillation method that divides the optimization process into blockwise optimization and joint optimization to address the limitation of layer-wise supervision and global supervision. Extensive experiments on ImageNet2012
(ResNet18/50, ShuffleNetV2, and MobileNetV2) demonstrate that the proposed approach can significantly improve the quantization model’s accuracy when only a few training samples are available. We further show that the method also extends to other computer vision architectures and tasks such as object detection.

Towards Low-bit Quantization of Deep Neural Networks with Limited Data

With the development of cheap image sensors, the amount of available image data have increased enormously, and the possibility of using crowdsourced collection methods has emerged. This calls for development of ways to handle all these data. In this paper, we present new tools that will enable efficient, flexible and robust map merging. Assuming that separate optimisations have been performed for the individual maps, we show how only relevant data can be stored in a low memory footprint representation. We use these representations to perform map merging so that the algorithm is invariant to the merging order and independent of the choice of coordinate system. The result is a robust algorithm that can be applied to several maps simultaneously. The result of a merge can also be represented with the same type of low-memory footprint format, which enables further merging and updating of the map in a hierarchical way. Furthermore, the method can perform loop closing and also detect changes in the scene between the capture of the different image sequences. Using both simulated and real data — from both a hand held mobile phone and from a drone — we verify the performance of the proposed method.

Generic Merging of Structure from Motion Maps with a Low Memory Footprint




Deep learning frameworks have been successfully applied to tackle many challenging tasks in pattern recognition and computer vision thanks to its ability to automatically extract representative features from the training data. Such type of data- driven approach, however, is subject to the criticism of too much dependency on the training set. In this research, we attempt to investigate the validity of this statement: ‘deep learning is only as good as its data’ by evaluating the performance of deep learning models using heterogeneous data sets, in which distinct representations of the same source data are employed for training/testing. We have examined three cases: low-resolution image, severely compressed input and halftone image in this work. Our preliminary results indicate that such dependency indeed exists. Classifier performance drops considerably when the model is tested with modified or transformed input. The best outcomes are obtained when the model is trained with hybrid input.

Investigation of DNN Model Robustness Using Heterogeneous Datasets


This paper presents a multi-channel Extended Depth of Field (EDF) method for digital cytology based on the stationary wavelet transform. With a coefficient selection rule adapted to a precise color recovery, a sharp image can be reconstructed even on images with transparent overlapping cells. The precision and the color fidelity of the proposed method is analyzed. Moreover, an experiment demonstrating the necessity of volume analysis in cytology to achieve precise segmentation on cell clumps is conducted, and the importance of color fidelity in this context is asserted. The proposed method was tested on pap-stained urothelial cells and gray-scale cervical cells with important overlapping.

Extended Depth of Field Preserving Color Fidelity For Automated Digital Cytology

Accurate and fine-grained information about the extent of damage to buildings is essential for directing Humanitarian Aid and Disaster Response (HADR) operations in the immediate aftermath of any natural calamity. In recent years, satellite and UAV (drone) imagery has been used for this purpose, sometimes aided by computer vision algorithms. Existing Computer Vision approaches for building damage assessment
typically rely on a two stage approach, consisting of building detection using an object detection model, followed by damage assessment through classification of the detected building tiles. These multi-stage methods are not end-to-end trainable, and
suffer from poor overall results. We propose RescueNet, a unified model that can simultaneously segment buildings and assess the damage levels to individual buildings and can be trained end-toend. In order to to model the composite nature of this problem, we propose a novel localization aware loss function, which consists
of a Binary Cross Entropy loss for building segmentation, and a foreground only selective Categorical Cross-Entropy loss for damage classification, and show significant improvement over the widely used Cross-Entropy loss. RescueNet is tested on the large scale and diverse xBD dataset and achieves significantly better building segmentation and damage classification performance than previous methods and achieves generalization across varied geographical regions and disaster types.

Premium content

Dual Path Multi-modal High-Order Features for Textual Content based Visual Question Answering

Downloads

Next from ICPR 2020

Neural Machine Registration for Motion Correction in Breast DCE-MRI

Similar lecture

Dual Path Multi-modal High-Order Features for Textual Content based Visual Question Answering