Italy


Current state-of-the-art approaches in the field of
Handwritten Text Recognition are predominately single task with
unigram, character level target units. In our work, we utilize a
Multi-task Learning scheme, training the model to perform de-
compositions of the target sequence with target units of different
granularity, from fine to coarse. We consider this method as
a way to utilize n-gram information, implicitly, in the training
process, while the final recognition is performed using only the
unigram output. Unigram decoding of such a multi-task approach
highlights the capability of the learned internal representations,
imposed by the different n-grams at the training step. We select
n-grams as our target units and we experiment from unigrams to
fourgrams, namely subword level granularities. These multiple
decompositions are learned from the network with task-specific
CTC losses. Concerning network architectures, we propose two
alternatives, namely the Hierarchical and the Block Multi-task.
Overall, our proposed model, even though evaluated only on the
unigram task outperforms its counterpart single-task by absolute
2.52% WER and 1.02% CER, in the greedy decoding, without
any computational overhead during inference, hinting towards
successfully imposing an implicit language model

ICPR 2020

Enhancing Handwritten Text Recognition with N-gram Sequence Decomposition and Multitask Learning

handwritten text recognition

multitask learning

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


In this paper we introduce a novel end-to-end framework for generation of large, aerial, city-scale, realistic synthetic image sequences with associated accurate and precise camera metadata. The two main purposes for this data are (i) to enable objective, quantitative evaluation of computer vision algorithms and methods such as feature detection, description, and matching or full computer vision pipelines such as 3D reconstruction; and (ii) to supply large amounts of high quality training data for deep learning guided computer vision methods. The proposed framework consists of three main modules, a 3D voxel renderer for data generation, a deep neural network for artifact removal, and a quantitative evaluation module for Multi-View Stereo (MVS) as an example. The 3D voxel renderer enables generation of seen or unseen views of a scene from arbitrary camera poses with accurate camera metadata parameters. The artifact removal module proposes a novel edge-augmented deep learning network with an explicit edgemap processing stream to remove image artifacts while preserving and recovering scene structures for more realistic results. Our experiments on two urban, city-scale, aerial datasets for Albuquerque (ABQ), NM and Los Angeles (LA), CA show promising results in terms of structural similarity to real data and accuracy of reconstructed 3D point clouds.

Deep Realistic Novel View Generation for City-Scale Aerial Images

Occlusion degrades the performance of human pose estimation. In this paper, we introduce targeted keypoint and body part occlusion attacks. The effects of the attacks are systematically analyzed on the best-performing methods. In addition, we propose occlusion specific data augmentation techniques against keypoint and part attacks. Our extensive experiments show that human pose estimation methods are not robust to occlusion and data augmentation does not solve the occlusion problems.

Tilting at windmills: Data augmentation for deep pose estimation does not help with occlusion


The prevalence of state-sponsored propaganda on the Internet has become a cause for concern in the recent years. While much effort has been made to identify state-sponsored Internet propaganda, the problem remains far from being solved because the ambiguous definition of propaganda leads to unreliable data labelling, and the huge amount of potential predictive features causes the models to be inexplicable. This paper is the first attempt to build a balanced dataset for this task. The dataset is comprised of propaganda by three different organizations across two time periods. A multi-model framework for detecting propaganda messages solely based on the visual and textual content is proposed which achieves a promising performance on detecting propaganda by the three organizations both for the same time period (training and testing on data from the same time period) (F1=0.869) and for different time periods (training on past, testing on future) (F1=0.697). To reduce the influence of false positive predictions, we change the threshold to test the relationship between the false positive and true positive rates and provide explanations for the predictions made by our models with visualization tools to enhance the interpretability of our framework. Our new dataset and general framework provide a strong benchmark for the task of identifying state-sponsored Internet propaganda and point out a potential path for future work on this task.

Multi-Modal Identification of State-Sponsored Propaganda on Social Media

I present an unsupervised approach that leverages raw aerial videos to learn to estimate planar homographic transformation between consecutive video frames. Previous learning-based estimators work on pairs of images to estimate their planar homographic transformations but suffer from severe over-fitting issues, especially when applying over aerial videos. To address this concern, we develop a sequential estimator that directly processes a sequence of video frames and estimates their pairwise planar homographic transformations in batches. We also incorporate a set of spatial-temporal knowledge to regularize the learning of such a sequence-to-sequence model. We collect a set of aerial videos and compare the proposed method to the alternative algorithms. Empirical studies suggest that our sequential model achieves significant improvement over alternative image-based methods and the knowledge-rich regularization further boosts our system performance. Our codes and dataset could be found at https://github.com/Paul-LiPu/DeepVideoHomography

Learning Knowledge-Rich Sequential Model for Planar Homography Estimation in Aerial Video

Principal Component Analysis (PCA) is a widely used technique for dimensionality reduction in various problem domains, including data compression, image processing, visualization, exploratory data analysis, pattern recognition, time-series prediction, and machine learning. Often, data is presented in a correlated paired manner such that there exist observable and correlated unobservable measurements. Unfortunately, traditional PCA techniques generally fail to optimally capture the leverageable correlations between such paired data as it does not yield a maximally correlated basis between the observable and unobservable counterparts. This instead is the objective of Canonical Correlation Analysis (and the more general Partial Least Squares methods); however, such techniques are still symmetric in maximizing correlation (covariance for PLSR) over all choices of the basis for both datasets without differentiating between observable and unobservable variables (except for the regression phase of PLSR). Further, these methods deviate from PCA's formulation objective to minimize approximation error, seeking instead to maximize correlation or covariance. While these are sensible optimization objectives, they are not equivalent to error minimization. We therefore introduce a new method of leveraging PCA between paired datasets in a dependently coupled manner, which is optimal with respect to approximation error during training. We generate a dependently coupled paired basis for which we relax orthogonality constraints in decomposing unreliable unobservable measurements. In doing so, this allows us to optimally capture the variations of the observable data while conditionally minimizing the expected prediction error for the unobservable component. We show preliminary results that demonstrate improved learning of our proposed method compared to that of traditional techniques.

Dependently Coupled Principal Component Analysis for Bivariate Inversion Problems

Predictive Coding Network (PCN) is an important neural network inspired by visual processing models in neuroscience. It combines the feedforward and feedback processing and has the architecture of recurrent neural networks (RNNs). This type of network is usually trained with backpropagation through time (BPTT). With infinite recurrent steps, PCN is a dynamic system. However, as one of the most important properties, stability is rarely studied in this type of network. Inspired by reservoir computing, we investigate the stability of hierarchical RNNs from the perspective of dynamic systems and propose a sufficient condition for their echo state property (ESP). Our study shows the global stability is determined by the stability of the local layers and the feedback between neighboring layers. Based on it, we further propose Weight Norm Supervision, a new algorithm that controls the stability of PCN dynamics by imposing different weight norm constraints on different parts of the network. We compare our approach with other training methods in terms of stability and prediction capability. The experiments show that our algorithm learns stable PCNs with a reliable prediction precision in the most effective and controllable way.

Learning Stable Deep Predictive Coding Networks with Weight Norm Supervision

Survival data is often collected in medical applications from a heterogeneous population of patients. While in the past, popular survival models focused on modeling the average effect of the covariates on survival outcomes, rapidly advancing sensing and information technologies have provided opportunities to further model the heterogeneity of the population as well as the non-linearity of the survival risk. With this motivation, we propose a new semi-parametric Bayesian Survival Rule List model in this paper. Our model derives a rule-based decision-making approach, while within the regime defined by each rule, survival risk is modelled via a Gaussian process latent variable model. Markov Chain Monte Carlo with a nested Laplace approximation on the Gaussian process posterior is used to search over the posterior of the rule lists efficiently. The use of ordered rule lists enables us to model heterogeneity while keeping the model complexity in check. Performance evaluations on a synthetic heterogeneous survival dataset and a real world sepsis survival dataset demonstrate the effectiveness of our model.

GPSRL: Learning Semi-Parametric Bayesian Survival Rule Lists from Heterogeneous Patient Data


The ubiquity of smart speakers is increasing, with a growing number of households utilising these devices to make calls over the telephony network. As the technology is typically configured to retain the cellular phone number of the user, it presents challenges in applications where knowledge of the true call origin is required. There are a wide range of makes and models for these devices, as is the case with cell phones, and it is challenging to detect the general category as a smart speaker or cell, independent of the designated phone number. In this paper, we present an approach to differentiate between calls originating from smart speakers and ones from cellular devices using only the audio. We present a feature set that characterises the relevant acoustic information, such as the degree of reverberation and noise, to distinguish between these categories. When evaluated on a dataset spanning multiple models for each device category, as well as different modes-of-usage and microphone-speaker distances, the method yields an Equal Error Rate (EER) of 12.6%.

Detection of Calls From Smart Speaker Devices


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.

Optimal Strategies For Comparing Covariates To Solve Matching Problems

The most existing studies in the facial age estimation assume training and test images are captured under similar shooting conditions. However, this is rarely valid in real-world applications, where training and test sets usually have different characteristics. In this paper, we advocate a cross-dataset protocol for age estimation benchmarking. In order to improve the cross-dataset age estimation performance, we mitigate the inherent bias caused by the learning algorithm itself.
To this end, we propose a novel loss function that is more effective for neural network training. The relative smoothness of the proposed loss function is its advantage with regards to the optimisation process performed by stochastic gradient descent (SGD). Compared with existing loss functions, the lower gradient of the proposed loss function leads to the convergence of SGD to a better optimum point, and consequently a better generalisation. The cross-dataset experimental results demonstrate the superiority of the proposed method over the state-of-the-art algorithms in terms of accuracy and generalisation capability.

Premium content

Downloads

Next from ICPR 2020

Deep Realistic Novel View Generation for City-Scale Aerial Images

Similar lecture

Watch Your Strokes: Improving Handwritten TextRecognition with Deformable Convolutions