Italy

The Softmax prediction function is widely used to train Deep Convolutional Neural Networks (DCNNs) for large-scale face recognition and other applications. The limitation of the softmax activation is that the resulting probability distribution always has a full support. This full support leads to larger intraclass variations. In this paper, we formulate a novel loss function, called Angular Sparsemax for face recognition. The proposed loss function promotes sparseness of the hypotheses prediction function similar to Sparsemax with Fenchel-Young regularisation. By introducing an additive angular margin on the score vector, the discriminatory power of the face embedding is further improved. The proposed loss function is experimentally validated on several databases in terms of recognition accuracy. Its performance compares well with the state of the art Arcface loss.


ICPR 2020

Anglar Sparsemax for Face Recognition

face recognition

loss function

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

The contemporary SLAM mapping systems assume a static environment and build a map that is then used for mobile robot navigation disregarding the dynamic changes in this environment. The paper at hand presents a novel solution for the problem of life-long mapping that continually updates a metric map represented as a 2D occupancy grid in large scale indoor environments with movable objects such as people, robots, objects etc. suitable for industrial applications. We formalize each cell’s occupancy as a failure analysis problem and contribute temporal persistence modeling (TPM), an algorithm for probabilistic prediction of the time that a cell in an observed location is expected to be “occupied” or “empty” given sparse prior observations from a task specific mobile robot. Our work is evaluated in Gazebo simulation environment against the nominal occupancy of cells and the estimated obstacles persistence. We also show that robot navigation with life-long mapping demands less re-plans and leads to more efficient navigation in highly dynamic environments.

Towards life-long mapping of dynamic environments using temporal persistence modeling

We propose a new approach to the problem of detecting out-of-distribution (OOD) inputs in image classifications with deep convolutional networks. We leverage so-called boundary samples to enforce low confidence (maximum softmax
probabilities) for inputs far away from the training data. In particular, we propose the boundary optimised samples (named BoS) training algorithm for generating them. At the end of the BoS training, all the boundary samples are in principle located on a specific level hypersurface with respect to the designed loss. Our contributions are i) the BoS training as an efficient alternative to generate boundary samples, ii) a robust algorithm therewith to enforce low confidence for OOD samples, and iii) experiments demonstrating improved OOD detection over the baseline.

Boundary Optimised Samples Training for Detecting Out-of-Distribution Images

Lung ultrasound imaging is reaching growing interest from the scientific community. On one side, thanks to its harmlessness and high descriptive power, this kind of diagnostic imaging has been largely adopted in sensitive applications, like the diagnosis and follow-up of preterm newborns in neonatal intensive care units. On the other side, state-of-the-art image analysis and pattern recognition approaches have recently proven their ability to fully exploit the rich information contained in these data, making them attractive for the research community. In this work, we present a thorough analysis of recent deep learning networks and training strategies carried out on a vast and challenging multicenter dataset comprising 87 patients with different diseases and gestational ages. These approaches are employed to assess the lung respiratory status from ultrasound images and are evaluated against a reference marker. The conducted analysis sheds some light on this problem by showing the critical points that can mislead the training procedure and proposes some adaptations to the specific data and task. The achieved results sensibly outperform those obtained by a previous work, which is based on textural features, and narrow the gap with the visual score predicted by the human experts.

Deep learning in the ultrasound evaluation of neonatal respiratory status

In this work, we propose a novel method for human action recognition that encodes efficiently the most discriminative appearance information of an action with explicit attention on representative pose features, into a new compact grid representation. Our GRAR (Grid-based Representation for Action Recognition) method is tested on several benchmark datasets demonstrating that our model can accurately recognize human actions, despite intra-class appearance variations and occlusion challenges.

A Grid-based Representation for Human Action Recognition

A face identification system compares an unknown input probe image to a gallery of labeled face images in order to determine the identity of the probe image. The result of identification is a ranked match list with the most similar gallery face image at the top (rank 1) and the least similar gallery face image at the bottom. In many systems, the top ranked gallery images may look very similar to the probe image as well as to each other and can sometimes result in the misidentification of the probe image. Such similar looking faces pertaining to different identities are referred to as lookalike faces. We hypothesize that a matcher specifically trained to disambiguate lookalike face images when combined with a regular face matcher will improve overall identification performance. This work proposes reranking the initial ranked match list using a disambiguator especially for lookalike face pairs. This work also evaluates schemes to select gallery images in the initial ranked match list that should be re-ranked. Experiments on the challenging TinyFace dataset shows that the proposed approach improves the closed-set identification accuracy of a state-of-the-art face matcher.

Lookalike Disambiguation: Improving Face Identification Performance at Top Ranks

Deep supervised hashing for image retrieval has attracted researchers' attention due to its high efficiency and superior retrieval performance. Most existing deep supervised hashing works, which are based on pairwise/triplet labels, suffer from the expensive computational cost and insufficient utilization of the semantics information. Recently, deep classwise hashing introduced a classwise loss supervised by class labels information alternatively; however, we find it still has its drawback. In this paper, we propose an improved deep classwise hashing, which enables hashing learning and class centers learning simultaneously. Specifically, we design a two-step strategy on center similarity learning. It interacts with the classwise loss to attract the class center to concentrate on the intra-class samples while pushing other class centers as far as possible. The centers similarity learning contributes to generating more compact and discriminative hashing codes. We conduct experiments on three benchmark datasets. It shows that the proposed method effectively surpasses the original method and outperforms state-of-the-art baselines under various commonly-used evaluation metrics for image retrieval.

Improved Deep Classwise Hashing With Centers Similarity Learning for Image Retrieval


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.

Bayesian Active Learning for Maximal Information Gain on Model Parameters

We propose a novel and efficient training method for RNNs by iteratively seeking a local minima on the loss surface within a small region, and leverage this directional vector for the update, in an outer-loop. We propose to utilize the Frank-Wolfe (FW) algorithm in this context. Although, FW implicitly involves normalized gradients, which can lead to a slow convergence rate, we develop a novel RNN training method that, surprisingly, even with the additional cost, the overall training cost is empirically observed to be lower than back-propagation. Our method leads to a new Frank-Wolfe method, that is in essence an SGD algorithm with a restart scheme. We prove that under certain conditions our algorithm has a sublinear convergence rate of $O(1/\epsilon)$ for $\epsilon$ error. We then conduct empirical experiments on several benchmark datasets including those that exhibit long-term dependencies, and show significant performance improvement. We also experiment with deep RNN architectures and show efficient training performance. Finally, we demonstrate that our training method is robust to noisy data.

RNN Training along Locally Optimal Trajectories via Frank-Wolfe Algorithm


Capturing images of documents is one of the easiest and most used methods of recording them. These images however, being captured with the help of handheld devices, often lead to undesirable distortions that are hard to remove. We propose a supervised Gated and Bifurcated Stacked U-Net module to predict a dewarping grid and create a distortion free image from the input. While the network is trained on synthetically warped document images, results are calculated on the basis of real world images. The novelty in our methods exists not only in a bifurcation of the U-Net to help eliminate the intermingling of the grid coordinates, but also in the use of a gated network which adds boundary and other minute line level details to the model. The end-to-end pipeline proposed by us achieves state-of-the-art performance on the DocUNet dataset after being trained on just 8 percent of the data used in previous methods.

A Gated and Bifurcated Stacked U-Net Module for Document Image Dewarping


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

Downloads

Next from ICPR 2020

Towards life-long mapping of dynamic environments using temporal persistence modeling

Similar lecture

Multi-level Deep Learning Vehicle Re-identification using Ranked-based Loss Functions