Abstract: In class-incremental semantic segmentation, we have no access to the labeled data of previous tasks. Therefore, when incrementally learning new classes, deep neural networks suffer from catastrophic forgetting of previously learned knowledge. To address this problem, we propose to apply a self-training approach that leverages unlabeled data, which is used for rehearsal of previous knowledge. Specifically, we first learn a temporary model for the current task, and then, pseudo labels for the unlabeled data are computed by fusing information from the old model of the previous task and the current temporary model. In addition, conflict reduction is proposed to resolve the conflicts of pseudo labels generated from both the old and temporary models. We show that maximizing self-entropy can further improve results by smoothing the overconfident predictions. Interestingly, in the experiments, we show that the auxiliary data can be different from the training data and that even general-purpose, but diverse auxiliary data can lead to large performance gains. The experiments demonstrate the state-of-the-art results: obtaining a relative gain of up to 114% on Pascal-VOC 2012 and 8.5% on the more challenging ADE20K compared to previous state-of-the-art methods.

Abstract: Recently, deep convolutional neural networks (CNNs) have provided outstanding performance in single image super-resolution (SISR). Despite their remarkable performance, the lack of high-frequency information in the recovered images remains a core problem. Moreover, as the networks increase in depth and width, deep CNN-based SR methods are faced with the challenge of computational complexity in practice. A promising and under-explored solution is to adapt the amount of compute based on the different frequency bands of the input. To this end, we present a novel Frequency-based Enhancement Block (FEB) which explicitly enhances the information of high frequencies while forwarding low-frequencies to the output. In particular, this block efficiently decomposes features into low- and high-frequency and assigns more computation to high-frequency ones. Thus, it can help the network generate more discriminative representations by explicitly recovering finer details. Our FEB design is simple and generic and can be used as a direct replacement of commonly used SR blocks with no need to change network architectures. We experimentally show that when replacing SR blocks with FEB we consistently improve the reconstruction error, while reducing the number of parameters in the model. Moreover, we propose a lightweight SR model — Frequency-based Enhancement Network (FENet) — based on FEB that matches the performance of larger models. Extensive experiments demonstrate that our proposal performs favorably against the state-of-the-art SR algorithms in terms of visual quality, memory footprint, and inference time. The code is available at https://github.com/pbehjatii/FENet

Abstract: Collaborative robotic industrial cells are workspaces where robots collaborate with human operators. In this context, safety is paramount, and for that a complete perception of the space where the collaborative robot is inserted is necessary. To ensure this, collaborative cells are equipped with a large set of sensors of multiple modalities, covering the entire work volume. However, the fusion of information from all these sensors requires an accurate extrinsic calibration. The calibration of such complex systems is challenging, due to the number of sensors and modalities, and also due to the small overlapping fields of view between the sensors, which are positioned to capture different viewpoints of the cell. This paper proposes a sensor to pattern methodology that can calibrate a complex system such as a collaborative cell in a single optimization procedure. Our methodology can tackle RGB and Depth cameras, as well as LiDARs. Results show that our methodology is able to accurately calibrate a collaborative cell containing three RGB cameras, a depth camera and three 3D LiDARs.

Abstract: Background and study aims Artificial intelligence is currently able to accurately predict the histology of colorectal polyps. However, systems developed to date use complex optical technologies and have not been tested in vivo. The objective of this study was to evaluate the efficacy of a new deep learning-based optical diagnosis system, ATENEA, in a real clinical setting using only high-definition white light endoscopy (WLE) and to compare its performance with endoscopists. Methods ATENEA was prospectively tested in real life on consecutive polyps detected in colorectal cancer screening colonoscopies at Hospital Clínic. No images were discarded, and only WLE was used. The in vivo ATENEA's prediction (adenoma vs non-adenoma) was compared with the prediction of four staff endoscopists without specific training in optical diagnosis for the study purposes. Endoscopists were blind to the ATENEA output. Histology was the gold standard. Results Ninety polyps (median size: 5 mm, range: 2-25) from 31 patients were included of which 69 (76.7 %) were adenomas. ATENEA correctly predicted the histology in 63 of 69 (91.3 %, 95 % CI: 82 %-97 %) adenomas and 12 of 21 (57.1 %, 95 % CI: 34 %-78 %) non-adenomas while endoscopists made correct predictions in 52 of 69 (75.4 %, 95 % CI: 60 %-85 %) and 20 of 21 (95.2 %, 95 % CI: 76 %-100 %), respectively. The global accuracy was 83.3 % (95 % CI: 74%-90 %) and 80 % (95 % CI: 70 %-88 %) for ATENEA and endoscopists, respectively. Conclusion ATENEA can accurately be used for in vivo characterization of colorectal polyps, enabling the endoscopist to make direct decisions. ATENEA showed a global accuracy similar to that of endoscopists despite an unsatisfactory performance for non-adenomatous lesions.

Abstract: The transcription of sheet music into some machine-readable format can be carried out manually. However, the complexity of music notation inevitably leads to burdensome software for music score editing, which makes the whole process
very time-consuming and prone to errors. Consequently, automatic transcription
systems for musical documents represent interesting tools.
Document analysis is the subject that deals with the extraction and processing
of documents through image and pattern recognition. It is a branch of computer
vision. Taking music scores as source, the field devoted to address this task is
known as Optical Music Recognition (OMR). Typically, an OMR system takes an
image of a music score and automatically extracts its content into some symbolic
structure such as MEI or MusicXML.
In this dissertation, we have investigated different methods for recognizing a
single staff section (e.g. scores for violin, flute, etc.), much in the same way as most text recognition research focuses on recognizing words appearing in a given line image. These methods are based in two different methodologies. On the one hand, we present two methods based on Recurrent Neural Networks, in particular, the
Long Short-Term Memory Neural Network. On the other hand, a method based on Sequence to Sequence models is detailed.
Music context is needed to improve the OMR results, just like language models
and dictionaries help in handwriting recognition. For example, syntactical rules
and grammars could be easily defined to cope with the ambiguities in the rhythm.
In music theory, for example, the time signature defines the amount of beats per
bar unit. Thus, in the second part of this dissertation, different methodologies
have been investigated to improve the OMR recognition. We have explored three
different methods: (a) a graphic tree-structure representation, Dendrograms, that
joins, at each level, its primitives following a set of rules, (b) the incorporation of Language Models to model the probability of a sequence of tokens, and (c) graph neural networks to analyze the music scores to avoid meaningless relationships between music primitives.
Finally, to train all these methodologies, and given the method-specificity of
the datasets in the literature, we have created four different music datasets. Two of them are synthetic with a modern or old handwritten appearance, whereas the
other two are real handwritten scores, being one of them modern and the other
old.

Abstract: Until recently, most computer vision models remained illiterate, largely ignoring the semantically rich and explicit information contained in scene text. Recent progress in scene text detection and recognition has recently allowed exploring its role in a diverse set of open computer vision problems, e.g. image classification, image-text retrieval, image captioning, and visual question answering to name a few. The explicit semantics of scene text closely requires specific modeling similar to language. However, scene text is a particular signal that has to be interpreted according to a comprehensive perspective that encapsulates all the visual cues in an image. Incorporating this information is a straightforward task for humans, but if we are unfamiliar with a language or scripture, achieving a complete world understanding is impossible (e.a. visiting a foreign country with a different alphabet). Despite the importance of scene text, modeling it requires considering the several ways in which scene text interacts with an image, processing and fusing an additional modality. In this thesis, we mainly focus
on two tasks, scene text-based fine-grained image classification, and cross-modal retrieval. In both studied tasks we identify existing limitations in current approaches and propose plausible solutions. Concretely, in each chapter: i) We define a compact way to embed scene text that generalizes to unseen words at training time while performing in real-time. ii) We incorporate the previously learned scene text embedding to create an image-level descriptor that overcomes optical character recognition (OCR) errors which is well-suited to the fine-grained image classification task. iii) We design a region-level reasoning network that learns the interaction through semantics among salient visual regions and scene text instances. iv) We employ scene text information in image-text matching and introduce the Scene Text Aware Cross-Modal retrieval StacMR task. We gather a dataset that incorporates scene text and design a model suited for the newly studied modality. v) We identify the drawbacks of current retrieval metrics in cross-modal retrieval. An image captioning metric is proposed as a way of better evaluating semantics in retrieved results. Ample experimentation shows that incorporating such semantics into a model yields better semantic results while
requiring significantly less data to converge.

Abstract: Accurate semantic segmentation models typically require significant computational resources, inhibiting their use in practical applications. Recent works rely on well-crafted lightweight models to achieve fast inference. However, these models cannot flexibly adapt to varying accuracy and efficiency requirements. In this paper, we propose a simple but effective slimmable semantic segmentation (SlimSeg) method, which can be executed at different capacities during inference depending on the desired accuracy-efficiency tradeoff. More specifically, we employ parametrized channel slimming by stepwise downward knowledge distillation during training. Motivated by the observation that the differences between segmentation results of each submodel are mainly near the semantic borders, we introduce an additional boundary guided semantic segmentation loss to further improve the performance of each submodel. We show that our proposed SlimSeg with various mainstream networks can produce flexible models that provide dynamic adjustment of computational cost and better performance than independent models. Extensive experiments on semantic segmentation benchmarks, Cityscapes and CamVid, demonstrate the generalization ability of our framework.