Italy

One of the significant challenges of deep neural networks is that the complex nature of the network prevents human comprehension of the outcome of the network. Consequently, the applicability of complex machine learning models is limited in the safety-critical domains, which incurs risk to life and property. To fully exploit the capabilities of complex neural networks, we propose a non-intrusive interpretability technique that uses the input and output of the model to generate a saliency map. The method works by empirically optimizing a randomly initialized input mask by localizing and weighing individual pixels according to their sensitivity towards the target class. Our experiments show that the proposed model interpretability approach performs better than the existing saliency map-based approaches methods at localizing the relevant input pixels. 

Furthermore, to obtain a global perspective on the target-specific explanation, we propose a saliency map reconstruction approach to generate acceptable variations of the salient inputs from the space of input data distribution for which the model outcome remains unaltered. Experiments show that our interpretability method can reconstruct the salient part of the input with a classification accuracy of 89%.

ICPR 2020

A generalizable saliency-map based interpretation of model outcome

images

interpretability

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 study attempts to use crowdsourcing to facilitate the operation of pattern-recognition-based video surveillance systems at an early stage. Target events (i.e. events to be detected during surveillance) are not frequently observed in recorded video, so achieving reliable surveillance on the basis of machine learning requires a sufficient amount of target data. Acquiring sufficient data is time-consuming. However, operating unreliable surveillance systems can induce many false alarms. Crowdsourcing is introduced to address this problem by verifying the unreliable results in data-driven surveillance. Experimental simulation conducted using monitoring video of Japanese black beef cattle demonstrates that crowdsourced verification successfully reduced false alarms in calving detection systems.

Crowdsourced Verification for Operating Calving Surveillance Systems at an Early Stage

This paper tackles on building a pattern recognition system that detects whether a pair of Japanese black beefs captured in a given image region is in a “mounting” action, which is known to be a sign critically important to be detected for cattle farmers before artificial insemination. The “mounting” action refers to a cattle’s action where a cow bends over another cow usually when either cow is in estrus. Although a pattern recognition-based approach for detecting such an action would be appreciated as being low-cost and robust, it had not been discussed much due to the complexity of the system architecture, unavailability of datasets, etc. This study presents i) our image dataset construction technique that exploits both object detection algorithm and crowdsourcing for collecting cattle pair images with labels of either “mounting” or not; and ii) a system for detecting the mounting action from any given image of a cattle pair, developed based on the dataset. Starting with an algorithm for extracting regions of cattle pairs from a video frame based on the intersection of single cattle regions, we then designed our crowdsourcing microtask in which crowd workers were given simple guidelines to annotate mounting-action-relevant labels to the extracted regions, to finally obtain a dataset. We also introduce our tandem-layered pattern recognition system trained with the dataset. The system is comprised of two serially-connected machine learning components, and is capable of more robustly detecting mounting actions even with a small amount of training data than a normal end-to-end neural network. Experimental comparisons demonstrated that our detection system was capable of detecting estrus with a precision rate of 80% and a recall rate of 76%.

Toward Building a Data-Driven System For Detecting Mounting Actions of Black Beef Cattle

Density regression has been widely employed in crowd counting. However, the frequency imbalance of pixel values in the density map is still an obstacle to improve the performance. In this paper, we propose a novel learning strategy for learning error-driven curriculum, which uses an additional network to supervise the training of the main network. A tutoring network called TutorNet is proposed to repetitively indicate the critical errors of the main network. TutorNet generates pixel-level weights to formulate the curriculum for the main network during training, so that the main network will assign a higher weight to those hard examples than easy examples. Furthermore, we scale the density map by a factor to enlarge the distance among inter-examples, which is well known to improve the performance. Extensive experiments on two challenging benchmark datasets show that our method has achieved state-of-the-art performance.

Learning Error-Driven Curriculum for Crowd Counting


State-of-the-art pedestrian detectors have achieved significant progress on non-occluded pedestrians, yet they are still struggling under heavy occlusions. The recent occlusion handling strategy of popular two-stage approaches is to build a two-branch architecture with the help of additional visible body annotations. Nonetheless, these methods still have some weaknesses. We propose in this paper a novel Mutual-Supervised Feature Modulation (MSFM) network, to better handle occluded pedestrian detection. To facilitate the MSFM module, we also propose a novel two-branch architecture, consisting of a standard full body detection branch and an extra visible body classification branch. To verify the effectiveness of our proposed method, extensive experiments are conducted on two challenging pedestrian datasets: Caltech and CityPersons, and our approach achieves superior performance compared to other state-of-the-art methods on both datasets, especially in heavy occlusion cases.

Mutual-Supervised Feature Modulation Network for Occluded Pedestrian Detection

Autism spectrum disorder (ASD) is a neuro- developmental disorder, which causes deficits in social lives. Early screening of ASD for young children is important to reduce the impact of ASD on people’s lives. Traditional screening methods mainly rely on protocol-based interviews and subjective evaluations from clinicians and domain experts, which requires advanced expertise and intensive labor. To standardize the process of ASD screening, we design a ”Responsive Social Smile” protocol and the associated experimental setup. Moreover, we propose a machine learning based assessment framework for early ASD screening. By integrating speech recognition and computer vision technologies, the proposed framework can quantitatively analyze children’s behaviors under well-designed protocols. We collect 196 stimulus samples from 41 children with an average age of 23.34 months, and the proposed method obtains 85.20% accuracy for predicting stimulus scores and 80.49% accuracy for the final ASD prediction. This result indicates that our model approaches the average level of domain experts in this ”Responsive Social Smile” protocol.

Responsive Social Smile: A Machine Learning based Multimodal Behavior Assessment Framework towards Early Stage Autism Screening


As scanners produce higher-resolution and more densely sampled images, this raises the challenge of data storage, transmission and communication within healthcare systems. Since the quality of medical images plays a crucial role in diagnosis accuracy, medical imaging compression techniques are desired to reduce scan bitrate while guaranteeing lossless reconstruction. This paper presents a lossless compression method that integrates a Recurrent Neural Network (RNN) as a 3D sequence prediction model. The aim is to learn the long dependencies of the voxel's neighbourhood in 3D using Long Short-Term Memory (LSTM) network then compress the residual error using arithmetic coding. Experiential results reveal that our method obtains a higher compression ratio achieving 15% saving compared to the state-of-the-art lossless compression standards, including JPEG-LS, JPEG2000, JP3D, HEVC, and PPMd. Our evaluation demonstrates that the proposed method generalizes well to unseen modalities CT and MRI for the lossless compression scheme. To the best of our knowledge, this is the first lossless compression method that uses LSTM neural network for 16-bit volumetric medical image compression.


MedZip: 3D Medical Images Lossless Compressor Using Recurrent Neural Network (LSTM)


Lossy image and video compression algorithms introduce several different types of visual artifacts that reduce the visual quality of the compressed media, and the higher the compression rate the higher is the strength of these artifacts. In this work, we describe an approach for visual quality improvement of compressed images and videos to be performed at presentation time, as to obtain the benefits of fast data transfer and reduced data storage, while enjoying a visual quality that could be obtained only reducing the compression rate. To obtain this result we propose to use a deep neural network trained using the NoGAN approach, adapting the popular DeOldify architecture used for colorization. We show how the proposed method can be applied both to image and video compression artifact removal and restoration.

A NoGAN approach for image and video restoration and compression artifact removal

Person re-identification (re-ID) which aims to recognize a pedestrian observed by non-overlapping cameras is a challenging task due to high variance between images from different viewpoints. Although remarkable progresses on research of re-ID had been obtained via leveraging the merits of deep learning framework through sufficient quantity training on a large amount of well labeled data, whereas, in real scenarios, re-ID generally suffers from lacking of well labeled training data. In this paper, we propose an attention-based model with attribute classification (AMAC) to facilitate a well trained model transferring across different data domains, which further enables an efficient cross-domain video-based person re-ID. Specifically, an attention-based sub-network is proposed for deep insight into the quality variations of local parts, hence, different local parts are cooperated with different weights to avoid the heavy occlusions or the cluttered background in datasets. Moreover, we introduce a transferred attribute classification sub-network to extract attribute-semantic features of any new target datasets without the requirement for new training attribute labels which are costly to annotate. Attribute-semantic features can be considered as valuable complementary information for person re-identification since they are robust to illumination varieties and different viewpoints across cameras. Due to the large gap between different datasets, we fine-tune each sub-network with pseudo labels on the target datasets respectively to strengthen the original model trained on other labeled datasets. Extensive comparable evaluations demonstrate the superiority of our AMAC in solving cross-domain person re-ID task on two benchmarks including PRID-2011 and iLIDS-VID.

Attention-based Model with Attribute Classification for Cross-domain Person Re-identification


This paper studies the problem of Person Re-Identification (ReID)
for large-scale applications. Recent research efforts have been devoted
to building complicated part models, which introduce considerably
high computational cost and memory consumption, inhibiting its practicability
in large-scale applications. This paper aims to develop a novel learning
strategy to find efficient feature embeddings while maintaining the
balance of accuracy and model complexity. More specifically, we find
by enhancing the classical triplet loss together with cross-entropy
loss, our method can explore the hard examples and build a discriminant
feature embedding yet compact enough for large-scale applications.
Our method is carried out progressively using Bayesian optimization,
and we call it the Progressive Learning Algorithm (PLA). Extensive
experiments on three large-scale datasets show that our PLA is comparable
or better than the state-of-the-arts. Especially, on the challenging
Market-1501 dataset, we achieve Rank-1=94.7%/mAP=89.4% while saving
at least 30\% parameters than strong part models.

Efficient Shadow Detection and Removal using Synthetic Data with Domain Adaptation

3D point cloud completion has been a long-standing challenge at scale, and corresponding per-point supervised training strategies suffered from cumbersome annotations. 2D supervision has recently emerged as a promising alternative for 3D tasks, but specific approaches for 3D point cloud completion still remain to be explored. To overcome these limitations, we propose an end-to-end method that directly lifts a single depth map to a completed point cloud. With one depth map as input, a multi-way novel depth view synthesis network (NDVNet) is designed to infer coarsely completed depth maps under various viewpoints. Meanwhile, a geometric depth perspective rendering module is introduced to utilize the raw input depth map to generate a re-projected depth map for each view. Therefore, the two parallelly generated depth maps for each view are further concatenated and refined by a depth completion network (DCNet). The final completed point cloud is fused from all refined depth views. Experimental results demonstrate the effectiveness of our proposed approach composed of the aforementioned components, to produce high-quality, state-of-the-art results on the popular SUNCG benchmark.

Premium content

A generalizable saliency-map based interpretation of model outcome

Downloads

Next from ICPR 2020

Crowdsourced Verification for Operating Calving Surveillance Systems at an Early Stage

Similar lecture

Top-DB-Net: Top DropBlock for Activation Enhancement in Person Re-Identification