Abstract: One direct consequence of recent advances in digital visual data generation and the direct availability of this information through the World-Wide Web, is a urgent demand for efficient image retrieval systems. The objective of image retrieval is to allow users to efficiently browse through this abundance of images. Due to the non-expert nature of the majority of the internet users, such systems should be user friendly, and therefore avoid complex user interfaces. In this chapter we investigate how high-level information provided by recently developed object recognition techniques can improve interactive image retrieval. Wel apply a bagof- word based image representation method to automatically classify images in a number of categories. These additional labels are then applied to improve the image retrieval system. Next to these high-level semantic labels, we also apply a low-level image description to describe the composition and color scheme of the scene. Both descriptions are incorporated in a user feedback image retrieval setting. The main objective is to show that automatic labeling of images with semantic labels can improve image retrieval results.

Abstract: One of the attractive characteristics of deep neural networks is their ability to transfer knowledge obtained in one domain to other related domains. As a result, high-quality networks can be trained in domains with relatively little training data. This property has been extensively studied for discriminative networks but has received significantly less attention for generative models. Given the often enormous effort required to train GANs, both computationally as well as in the dataset collection, the re-use of pretrained GANs is a desirable objective. We propose a novel knowledge transfer method for generative models based on mining the knowledge that is most beneficial to a specific target domain, either from a single or multiple pretrained GANs. This is done using a miner network that identifies which part of the generative distribution of each pretrained GAN outputs samples closest to the target domain. Mining effectively steers GAN sampling towards suitable regions of the latent space, which facilitates the posterior finetuning and avoids pathologies of other methods such as mode collapse and lack of flexibility. We perform experiments on several complex datasets using various GAN architectures (BigGAN, Progressive GAN) and show that the proposed method, called MineGAN, effectively transfers knowledge to domains with few target images, outperforming existing methods. In addition, MineGAN can successfully transfer knowledge from multiple pretrained GANs.

Abstract: One of the challenges in computer vision models, especially sign language, is real-time recognition. In this work, we present a simple yet low-complex and efficient model, comprising single shot detector, 2D convolutional neural network, singular value decomposition (SVD), and long short term memory, to real-time isolated hand sign language recognition (IHSLR) from RGB video. We employ the SVD method as an efficient, compact, and discriminative feature extractor from the estimated 3D hand keypoints coordinators. Despite the previous works that employ the estimated 3D hand keypoints coordinates as raw features, we propose a novel and revolutionary way to apply the SVD to the estimated 3D hand keypoints coordinates to get more discriminative features. SVD method is also applied to the geometric relations between the consecutive segments of each finger in each hand and also the angles between these sections. We perform a detailed analysis of recognition time and accuracy. One of our contributions is that this is the first time that the SVD method is applied to the hand pose parameters. Results on four datasets, RKS-PERSIANSIGN (99.5±0.04), First-Person (91±0.06), ASVID (93±0.05), and isoGD (86.1±0.04), confirm the efficiency of our method in both accuracy (mean+std) and time recognition. Furthermore, our model outperforms or gets competitive results with the state-of-the-art alternatives in IHSLR and hand action recognition.

Abstract: One of the fundamental problems of computer vision is to represent images with compact semantically relevant embeddings. These embeddings could then be used in a wide variety of applications, such as image retrieval, object detection, and video search. The main objective of this thesis is to study image embeddings from two aspects: color embeddings and deep embeddings.
In the first part of the thesis we start from hand-crafted color embeddings. We propose a method to order the additional color names according to their complementary nature with the basic eleven color names. This allows us to compute color name representations with high discriminative power of arbitrary length. Psychophysical experiments confirm that our proposed method outperforms baseline approaches. Secondly, we learn deep color embeddings from weakly labeled data by adding an attention strategy. The attention branch is able to correctly identify the relevant regions for each class. The advantage of our approach is that it can learn color names for specific domains for which no pixel-wise labels exists.
In the second part of the thesis, we focus on deep embeddings. Firstly, we address the problem of compressing large embedding networks into small networks, while maintaining similar performance. We propose to distillate the metrics from a teacher network to a student network. Two new losses are introduced to model the communication of a deep teacher network to a small student network: one based on an absolute teacher, where the student aims to produce the same embeddings as the teacher, and one based on a relative teacher, where the distances between pairs of data points is communicated from the teacher to the student. In addition, various aspects of distillation have been investigated for embeddings, including hint and attention layers, semi-supervised learning and cross quality distillation. Finally, another aspect of deep metric learning, namely lifelong learning, is studied. We observed some drift occurs during training of new tasks for metric learning. A method to estimate the semantic drift based on the drift which is experienced by data of the current task during its training is introduced. Having this estimation, previous tasks can be compensated for this drift, thereby improving their performance. Furthermore, we show that embedding networks suffer significantly less from catastrophic forgetting compared to classification networks when learning new tasks.

Abstract: One of the main characteristics of agricultural fields is that the appearance of different crops and their growth status, in an aerial image, is varied, and has a wide range of radiometric values and high level of variability. The extraction of these fields and their monitoring are activities that require a high level of human intervention. In this article, we propose a novel automatic algorithm, named deep network energy-minimization (DeepNEM), to extract agricultural fields in aerial images. The model-guided process selects the most relevant image clues extracted by a deep network, completes them and finally generates regions that represent the agricultural fields under a minimization scheme. DeepNEM has been tested over a broad range of fields in terms of size, shape, and content. Different measures were used to compare the DeepNEM with other methods, and to prove that it represents an improved approach to achieve a high-quality segmentation of agricultural fields. Furthermore, this article also presents a new public dataset composed of 1200 images with their parcels boundaries annotations.

Abstract: One of the major prerequisites for any deep learning approach is the availability of large-scale training data. When dealing with scanned document images in real world scenarios, the principal information of its content is stored in the layout itself. In this work, we have proposed an automated deep generative model using Graph Neural Networks (GNNs) to generate synthetic data with highly variable and plausible document layouts that can be used to train document interpretation systems, in this case, specially in digital mailroom applications. It is also the first graph-based approach for document layout generation task experimented on administrative document images, in this case, invoices.

Abstract: One of the most widely used assumptions in supervised learning is that data is independent and identically distributed. This assumption does not hold true in many real cases. Sequential learning is the discipline of machine learning that deals with dependent data such that neighboring examples exhibit some kind of relationship. In the literature, there are different approaches that try to capture and exploit this correlation, by means of different methodologies. In this paper we focus on meta-learning strategies and, in particular, the stacked sequential learning approach. The main contribution of this work is two-fold: first, we generalize the stacked sequential learning. This generalization reflects the key role of neighboring interactions modeling. Second, we propose an effective and efficient way of capturing and exploiting sequential correlations that takes into account long-range interactions by means of a multi-scale pyramidal decomposition of the predicted labels. Additionally, this new method subsumes the standard stacked sequential learning approach. We tested the proposed method on two different classification tasks: text lines classification in a FAQ data set and image classification. Results on these tasks clearly show that our approach outperforms the standard stacked sequential learning. Moreover, we show that the proposed method allows to control the trade-off between the detail and the desired range of the interactions.

Abstract: Online continual learning aims to get closer to a live learning experience by learning directly on a stream of data with temporally shifting distribution and by storing a minimum amount of data from that stream. In this empirical evaluation, we evaluate various methods from the literature that tackle online continual learning. More specifically, we focus on the class-incremental setting in the context of image classification, where the learner must learn new classes incrementally from a stream of data. We compare these methods on the Split-CIFAR100 and Split-TinyImagenet benchmarks, and measure their average accuracy, forgetting, stability, and quality of the representations, to evaluate various aspects of the algorithm at the end but also during the whole training period. We find that most methods suffer from stability and underfitting issues. However, the learned representations are comparable to i.i.d. training under the same computational budget. No clear winner emerges from the results and basic experience replay, when properly tuned and implemented, is a very strong baseline. We release our modular and extensible codebase at this https URL based on the avalanche framework to reproduce our results and encourage future research.