Deep Learning (DL) is a two-step classification model that consists feature learning, generating feature representations using unsupervised ways and the supervised learning stage at the last step of model using at least two hidden layers on the proposed structures by fully connected layers depending on of the artificial neural networks. The optimization of the predefined classification parameters for the supervised models eases reaching the global optimality with exact zero training error. The autoencoder (AE) models are the highly generalized ways of the unsupervised stages for the DL to define the output weights of the hidden neurons with various representations. As alternatively to the conventional Extreme Learning Machines (ELM) AE, Hessenberg decomposition-based ELM autoencoder (HessELM-AE) is a novel kernel to generate different presentations of the input data within the intended sizes of the models. The aim of the study is analyzing the performance of the novel Deep AE kernel for clinical availability on electroencephalogram (EEG) with stroke patients. The slow cortical potentials (SCP) training in stroke patients during eight neurofeedback sessions were analyzed using Hilbert-Huang Transform. The statistical features of different frequency modulations were fed into the Deep ELM model for generative AE kernels. The novel Deep ELM-AE kernels have discriminated the brain activity with high classification performances for positivity and negativity tasks in stroke patients.

In this period where the number of systems developed by utilizing unmanned aerial technology is increasing day by day, unmanned aerial vehicles will be used to achieve the targets of minimizing the destruction of our forests which are the lungs of the world and optimizing the usage of workforce and time resources [1, 2, 3]. As a result of the application carried out in line with the subject of this paper, it is proposed that the system based on the detection of smoke image with unmanned aerial vehicle can provide a great benefit in reducing the error rate occurring in fire detection. The microprocessor in the system has been trained with deep learning methods and has been given the ability to recognize smoke image, which is the earliest sign of fire diagnosis. The most fundamental problem in the common algorithms used in fire detection is the high level of false alarm and overlook rate [4,5].Confirming the result obtained from the detection and defining an additional proof will increase the reliability of the system as well as the accuracy. Since the drones provide a mobile vision, the point of view can be controlled by the ground station can manipulate it for the sake of the accuracy of the result. The application developed in line with the subject of the paper was implemented in both simulation and physical environments and the advantages of early fire detection system and analysis results are discussed in the conclusion section of the article.

Large pre-trained multilingual models like mBERT, XLM-R achieve state of the art results on language understanding tasks. However, they are not well suited for latency critical applications on both servers and edge devices. It's important to reduce the memory and compute resources required by these models. To this end, we propose pQRNN, a projection-based embedding-free neural encoder that is tiny and effective for natural language processing tasks. Without pre-training, pQRNNs significantly outperform LSTM models with pre-trained embeddings despite being 140x smaller. With the same number of parameters, they outperform transformer baselines thereby showcasing their parameter efficiency. Additionally, we show that pQRNNs are effective student architectures for distilling large pre-trained language models. We perform careful ablations which study the effect of pQRNN parameters, data augmentation, and distillation settings. On MTOP, a challenging multilingual semantic parsing dataset, pQRNN students achieve 95.9\% of the performance of an mBERT teacher while being 350x smaller. On mATIS, a popular parsing task, pQRNN students on average are able to get to 97.1\% of the teacher while again being 350x smaller. Our strong results suggest that our approach is great for latency-sensitive applications while being able to leverage large mBERT-like models.

This thesis is a proof of concept for the potential of Variational Auto-Encoder (VAE) on representation learning of real-world Knowledge Graphs (KG). Inspired by successful approaches to the generation of molecular graphs, we evaluate the capabilities of our model, the Relational Graph Variational Auto-Encoder (RGVAE). The impact of the modular hyperparameter choices, encoding through graph convolutions, graph matching and latent space prior, is compared. The RGVAE is first evaluated on link prediction. The mean reciprocal rank (MRR) scores on the two datasets FB15K-237 and WN18RR are compared to the embedding-based model DistMult. A variational DistMult and a RGVAE without latent space prior constraint are implemented as control models. The results show that between different settings, the RGVAE with relaxed latent space, scores highest on both datasets, yet does not outperform the DistMult. Further, we investigate the latent space in a twofold experiment: first, linear interpolation between the latent representation of two triples, then the exploration of each latent dimension in a $95\%$ confidence interval. Both interpolations show that the RGVAE learns to reconstruct the adjacency matrix but fails to disentangle. For the last experiment we introduce a new validation method for the FB15K-237 data set. The relation type-constrains of generated triples are filtered and matched with entity types. The observed rate of valid generated triples is insignificantly higher than the random threshold. All generated and valid triples are unseen. A comparison between different latent space priors, using the $\delta$-VAE method, reveals a decoder collapse. Finally we analyze the limiting factors of our approach compared to molecule generation and propose solutions for the decoder collapse and successful representation learning of multi-relational KGs.

The significant power of deep learning networks has led to enormous development in object detection. Over the last few years, object detector frameworks have achieved tremendous success in both accuracy and efficiency. However, their ability is far from that of human beings due to several factors, occlusion being one of them. Since occlusion can happen in various locations, scale, and ratio, it is very difficult to handle. In this paper, we address the challenges in occlusion handling in generic object detection in both outdoor and indoor scenes, then we refer to the recent works that have been carried out to overcome these challenges. Finally, we discuss some possible future directions of research.

Data annotation plays a crucial role in ensuring your named entity recognition (NER) projects are trained with the right information to learn from. Producing the most accurate labels is a challenge due to the complexity involved with annotation. Label inconsistency between multiple subsets of data annotation (e.g., training set and test set, or multiple training subsets) is an indicator of label mistakes. In this work, we present an empirical method to explore the relationship between label (in-)consistency and NER model performance. It can be used to validate the label consistency (or catches the inconsistency) in multiple sets of NER data annotation. In experiments, our method identified the label inconsistency of test data in SCIERC and CoNLL03 datasets (with 26.7% and 5.4% label mistakes). It validated the consistency in the corrected version of both datasets.

Deep learning models are susceptible to adversarial examples that have imperceptible perturbations in the original input, resulting in adversarial attacks against these models. Analysis of these attacks on the state of the art transformers in NLP can help improve the robustness of these models against such adversarial inputs. In this paper, we present Adv-OLM, a black-box attack method that adapts the idea of Occlusion and Language Models (OLM) to the current state of the art attack methods. OLM is used to rank words of a sentence, which are later substituted using word replacement strategies. We experimentally show that our approach outperforms other attack methods for several text classification tasks.

We present a plug-in replacement for batch normalization (BN) called exponential moving average normalization (EMAN), which improves the performance of existing student-teacher based self- and semi-supervised learning techniques. Unlike the standard BN, where the statistics are computed within each batch, EMAN, used in the teacher, updates its statistics by exponential moving average from the BN statistics of the student. This design reduces the intrinsic cross-sample dependency of BN and enhance the generalization of the teacher. EMAN improves strong baselines for self-supervised learning by 4-6/1-2 points and semi-supervised learning by about 7/2 points, when 1%/10% supervised labels are available on ImageNet. These improvements are consistent across methods, network architectures, training duration, and datasets, demonstrating the general effectiveness of this technique.

We study the robustness of reinforcement learning (RL) with adversarially perturbed state observations, which aligns with the setting of many adversarial attacks to deep reinforcement learning (DRL) and is also important for rolling out real-world RL agent under unpredictable sensing noise. With a fixed agent policy, we demonstrate that an optimal adversary to perturb state observations can be found, which is guaranteed to obtain the worst case agent reward. For DRL settings, this leads to a novel empirical adversarial attack to RL agents via a learned adversary that is much stronger than previous ones. To enhance the robustness of an agent, we propose a framework of alternating training with learned adversaries (ATLA), which trains an adversary online together with the agent using policy gradient following the optimal adversarial attack framework. Additionally, inspired by the analysis of state-adversarial Markov decision process (SA-MDP), we show that past states and actions (history) can be useful for learning a robust agent, and we empirically find a LSTM based policy can be more robust under adversaries. Empirical evaluations on a few continuous control environments show that ATLA achieves state-of-the-art performance under strong adversaries. Our code is available at https://github.com/huanzhang12/ATLA_robust_RL.

This observation--that to understand Proust's text requires knowledge of various kinds--is not a new one. We came across it before, in the context of the Cyc project. Remember that Cyc was supposed to be given knowledge corresponding to the whole of consensus reality, and the Cyc hypothesis was that this would yield human-level general intelligence. Researchers in knowledge-based AI would be keen for me to point out to you that, decades ago, they anticipated exactly this issue. But it is not obvious that just continuing to refine deep learning techniques will address this problem.