Italy

Multi-scale feature fusion has been an effective way for improving the performance of semantic segmentation. How- ever, current methods generally fail to consider the semantic gaps between the shallow (low-level) and deep (high-level) features and thus the fusion methods may not be optimal. In this paper, to address the issues of the semantic gap between the feature from different layers, we propose a unified framework based on the U-shape encoder-decoder architecture, named Enhanced Feature Pyramid Network (EFPN). Specifically, the semantic enhancement module (SEM), edge extraction module (EEM), and context aggregation model (CAM) are incorporated into the decoder network to improve the robustness of the multi- level features aggregation. In addition, a global fusion model (GFM), which in the encoder branch is proposed to capture more semantic information in the deep layers and effectively transmit the high-level semantic features to each layer. Extensive experiments are conducted and the results show that the proposed framework achieves the state-of-the-art results on three public datasets, namely PASCAL VOC 2012, Cityscapes, and PASCAL Context. Furthermore, we also demonstrate that the proposed method is effective for other visual tasks that require frequent fusing features and upsampling.

ICPR 2020

Enhanced Feature Pyramid Network for Semantic Segmentation

sence parse

semantic segmentation

deep learning

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


This presentation proposes a method that can classify foot strike types using a deep learning model and can extract unknown factors, which enables to evaluate running motions without being influenced by biases of sports experts, using the contribution degree of input values (CDIV). Accelerometers are attached to the runner’s body, and when the runner runs, a fixed camera observes the runner and acquires a video sequence synchronously with the accelerometers. To train a deep learning model for classifying foot strikes, we annotate foot strike acceleration data for RFS (Rearfoot strike) or non-RFS objectively by watching the video. To interpret the unknown factors extracted from the learned model, we calculate two CDIVs: the contributions of the resampling time and the accelerometer value to the output (foot strike type) . 
Experiments on classifying unknown runners’ foot strikes were conducted. As a common result to sport science, it is confirmed that the CDIVs contribute highly at the time of the right foot strike, and the sensor values corresponding to the right and left tibias contribute highly to classifying the foot strikes. Experimental results show the right tibia is important for classifying foot strikes. This is because many of the training data represent difference between the two foot strikes in the right tibia. As a conclusion, our proposed method could extract unknown factors from the classifier and could interpret the factors that contain similar knowledge to the prior knowledge of experts, as well as new findings that are not included in conventional knowledge.

Extracting and Interpreting Unknown Factors with Classifier for Foot Strike Types in Running

This paper addresses unsupervised few-shot object recognition, where all training images are unlabeled, and test images are divided into queries and a few labeled support images per object class of interest. The training and test images do not share object classes. We extend the vanilla GAN with two loss functions, both aimed at self-supervised learning. The ﬁrst is a reconstruction loss that enforces the discriminator to reconstruct the probabilistically sampled latent code which has been used for generating the “fake” image. The second is a triplet loss that enforces the discriminator to output image encodings that are closer for more similar images. Evaluation, comparisons, and detailed ablation studies are done in the context of few-shot classiﬁcation. Our approach signiﬁcantly outperforms the state of the art on the Mini-Imagenet and Tiered-Imagenet datasets.

A Self-supervised GAN for Unsupervised Few-shot Object Recognition


Optical Character Recognition (OCR) is booming in recent years. As we all know, the text recognition results are vulnerable to slight perturbation in input images, thus a method for measuring how reliable the results are is crucial. 
In this paper, We propose a coarse-to-fine method for lexicon-free OCR confidence prediction which can be embedded with any text sequence recognition networks.

Robust Lexicon-Free Confidence Prediction for Text Recognition


There have been attempts to solve the Single Image Super-resolution problem using lightweight networks, considering limited computational resources for real-world applications. In previous works, residual paths in networks are pre-fixed and manually designed by human researchers. The proposed method allows the network to dynamically select residual paths depending on the input image, based on the idea of attention mechanism.

Single Image Super-Resolution with Dynamic Residual Attention

Object detection has been paid rising attention in computer vision field. Convolutional Neural Networks (CNNs) extract high-level semantic features of images, which directly determine the performance of object detection. As a common solution, embedding integration modules into CNNs can enrich extracted features and thereby improve the performance. However, the instability and inconsistency of internal multiple branches exist in these modules. To address this problem, we propose a novel multibranch module called Efficient-Receptive Field Block (E-RFB), in which multiple levels of features are combined for network optimization. Specifically, by downsampling and increasing depth, the E-RFB provides sufficient RF. Second, in order to eliminate the inconsistency across different branches, a novel spatial attention mechanism, namely, Group Spatial Attention Module (GSAM) is proposed. The GSAM gradually narrows a feature map by channel grouping; thus it encodes the information between spatial and channel dimensions into the final attention heat map. Third, the proposed module can be easily joined in various CNNs to enhance feature representation as a plug-and-play component. With SSD-style detectors, our method halves the parameters of the original detection head and achieves high accuracy on the PASCAL VOC and MS COCO datasets. Moreover, the proposed method achieves superior performance compared with state-of-the-art methods based on similar framework.


Efficient-Receptive Field Block with Group Spatial Attention Mechanism for Object Detection

This paper presents a new system for predicting the spatial change of Nasopharyngeal carcinoma(NPC) and organ-at-risks (OARs) volumes over the course of the radiation therapy (RT) treatment for facilitating the workflow of adaptive radiation therapy (ART).

A Deep Learning-Based Method for Predicting Volumes of Nasopharyngeal Carcinoma for Adaptive Radiation Therapy Treatment

Domain knowledge is useful to improve the generalization performance of learning machines. Sign constraints are a handy representation to combine domain knowledge with learning machine. In this poster presentation, we consider constraining the signs of the weight coefficients in learning the linear support vector machine, and develop two optimization algorithms for minimizing the empirical risk under the sign constraints. One of the two algorithms is based on the projected gradient method. The second algorithm is based on the Frank-Wolfe method that also converges sublinearly and possesses a clear termination criterion. Finally, we empirically demonstrate that the sign constraints are a promising technique when similarities to the training examples compose the feature vector.

Learning Sign-Constrained Support Vector Machines

LiDAR based 3D object detectors typically need a large amount of detailed-labeled point cloud data for training, but these detailed labels are commonly expensive to acquire. In this paper, we propose a manual-label free 3D detection algorithm that leverages the CARLA simulator to generate a large amount of self-labeled training samples and introduces a novel Domain Adaptive VoxelNet (DA-VoxelNet) that can cross the distribution gap from the synthetic data to the real scenario. The self-labeled training samples are generated by a set of high quality 3D models embedded in a CARLA simulator and a proposed LiDAR-guided sampling algorithm. Then a DA-VoxelNet that integrates both a sample-level DA module and an anchor-level DA module is proposed to enable the detector trained by the synthetic data to adapt to real scenario. Experimental results show that the proposed unsupervised DA 3D detector on KITTI evaluation set can achieve 76.66% and 56.64% mAP on BEV mode and 3D mode respectively. The results reveal a promising perspective of training a LIDAR-based 3D detector without any hand-tagged label.

Manual-Label Free 3D Detection via An Open-Source Simulator

Most object detection methods focus only on the local information near the region proposal and ignore the object's global semantic relation and local spatial relation information, resulting in limited performance. To capture and explore these important relations, we propose a detection method based on a graph convolutional network (GCN). Two independent relation graph networks are used to obtain the global semantic information of the object in labels and the local spatial information in images. Semantic relation networks can implicitly acquire global knowledge, and by constructing a directed graph on the dataset, each node is represented by the word embedding of labels and then sent to the GCN to obtain high-level semantic representation. The spatial relation network encodes the relation by the positional relation module and the visual connection module, and enriches the object features through local key information from objects. The feature representation is further improved by aggregating the outputs of the two networks. Instead of directly disseminating visual features in the network, the dual-graph network explores more advanced feature information, giving the detector the ability to obtain key relations in labels and region proposals. Experiments on the PASCAL VOC and MS COCO datasets demonstrate that key relation information significantly improve the performance of detection with better ability to detect small objects and reasonable bounding box. The results on COCO dataset demonstrate our method obtains around 32.3% improvement on AP in terms of small objects.

Object Detection Using Dual Graph Network

The activities we do in our daily-life are generally carried out as a succession of atomic actions, following a logical order.
During a video sequence, actions usually follow a logical order.
In this paper, we propose a hybrid approach resulting from the fusion of a deep learning neural network with a Bayesian-based approach. The latter models human-object interactions and transition between actions. The key idea is to combine both approaches in the final prediction.
We validate our strategy in two public datasets: CAD-120 and Watch-n-Patch. We show that our fusion approach yields performance gains in accuracy of respectively +4 percentage points (pp) and +6 pp over a baseline approach. Temporal action recognition performances are clearly improved by the fusion, especially when classes are imbalanced.

Premium content

Enhanced Feature Pyramid Network for Semantic Segmentation

Downloads

Next from ICPR 2020

Extracting and Interpreting Unknown Factors with Classifier for Foot Strike Types in Running

Similar lecture

Unsupervised Domain Adaptation with Multiple Domain Discriminators and Adaptive Self-Training