The majority of human deaths and injuries are caused by traffic accidents. A million people worldwide die each year due to traffic accident injuries, consistent with the World Health Organization. Drivers who do not receive enough sleep, rest, or who feel weary may fall asleep behind the wheel, endangering both themselves and other road users. The research on road accidents specified that major road accidents occur due to drowsiness while driving. These days, it is observed that tired driving is the main reason to occur drowsiness. Now, drowsiness becomes the main principle for to increase in the number of road accidents. This becomes a major issue in a world which is very important to resolve as soon as possible. The predominant goal of all devices is to improve the performance to detect drowsiness in real time. Many devices were developed to detect drowsiness, which depend on different artificial intelligence algorithms. So, our research is also related to driver drowsiness detection which can identify the drowsiness of a driver by identifying the face and then followed by eye tracking. The extracted eye image is matched with the dataset by the system. With the help of the dataset, the system detected that if eyes were close for a certain range, it could ring an alarm to alert the driver and if the eyes were open after the alert, then it could continue tracking. If the eyes were open then the score that we set decreased and if the eyes were closed then the score increased. This paper focus to resolve the problem of drowsiness detection with an accuracy of 80% and helps to reduce road accidents.
Recent development of neural vocoders based on the generative adversarial neural network (GAN) has shown their advantages of generating raw waveform conditioned on mel-spectrogram with fast inference speed and lightweight networks. Whereas, it is still challenging to train a universal neural vocoder that can synthesize high-fidelity speech from various scenarios with unseen speakers, languages, and speaking styles. In this paper, we propose DSPGAN, a GAN-based universal vocoder for high-fidelity speech synthesis by applying the time-frequency domain supervision from digital signal processing (DSP). To eliminate the mismatch problem caused by the ground-truth spectrograms in training phase and the predicted spectrograms in inference phase, we leverage the mel-spectrogram extracted from the waveform generated by a DSP module, rather than the predicted mel-spectrogram from the Text-to-Speech (TTS) acoustic model, as the time-frequency domain supervision to the GAN-based vocoder. We also utilize sine excitation as the time-domain supervision to improve the harmonic modeling and eliminate various artifacts of the GAN-based vocoder. Experimental results show that DSPGAN significantly outperforms the compared approaches and can generate high-fidelity speech based on diverse data in TTS.
Recent works have shown that self-supervised learning can achieve remarkable robustness when integrated with adversarial training (AT). However, the robustness gap between supervised AT (sup-AT) and self-supervised AT (self-AT) remains significant. Motivated by this observation, we revisit existing self-AT methods and discover an inherent dilemma that affects self-AT robustness: either strong or weak data augmentations are harmful to self-AT, and a medium strength is insufficient to bridge the gap. To resolve this dilemma, we propose a simple remedy named DYNACL (Dynamic Adversarial Contrastive Learning). In particular, we propose an augmentation schedule that gradually anneals from a strong augmentation to a weak one to benefit from both extreme cases. Besides, we adopt a fast post-processing stage for adapting it to downstream tasks. Through extensive experiments, we show that DYNACL can improve state-of-the-art self-AT robustness by 8.84% under Auto-Attack on the CIFAR-10 dataset, and can even outperform vanilla supervised adversarial training for the first time. Our code is available at \url{https://github.com/PKU-ML/DYNACL}.
Audio source separation is often achieved by estimating the magnitude spectrogram of each source, and then applying a phase recovery (or spectrogram inversion) algorithm to retrieve time-domain signals. Typically, spectrogram inversion is treated as an optimization problem involving one or several terms in order to promote estimates that comply with a consistency property, a mixing constraint, and/or a target magnitude objective. Nonetheless, it is still unclear which set of constraints and problem formulation is the most appropriate in practice. In this paper, we design a general framework for deriving spectrogram inversion algorithm, which is based on formulating optimization problems by combining these objectives either as soft penalties or hard constraints. We solve these by means of algorithms that perform alternating projections on the subsets corresponding to each objective/constraint. Our framework encompasses existing techniques from the literature as well as novel algorithms. We investigate the potential of these approaches for a speech enhancement task. In particular, one of our novel algorithms outperforms other approaches in a realistic setting where the magnitudes are estimated beforehand using a neural network.
Miniaturized autonomous unmanned aerial vehicles (UAVs) are an emerging and trending topic. With their form factor as big as the palm of one hand, they can reach spots otherwise inaccessible to bigger robots and safely operate in human surroundings. The simple electronics aboard such robots (sub-100mW) make them particularly cheap and attractive but pose significant challenges in enabling onboard sophisticated intelligence. In this work, we leverage a novel neural architecture search (NAS) technique to automatically identify several Pareto-optimal convolutional neural networks (CNNs) for a visual pose estimation task. Our work demonstrates how real-life and field-tested robotics applications can concretely leverage NAS technologies to automatically and efficiently optimize CNNs for the specific hardware constraints of small UAVs. We deploy several NAS-optimized CNNs and run them in closed-loop aboard a 27-g Crazyflie nano-UAV equipped with a parallel ultra-low power System-on-Chip. Our results improve the State-of-the-Art by reducing the in-field control error of 32% while achieving a real-time onboard inference-rate of ~10Hz@10mW and ~50Hz@90mW.
Next POI recommendation intends to forecast users' immediate future movements given their current status and historical information, yielding great values for both users and service providers. However, this problem is perceptibly complex because various data trends need to be considered together. This includes the spatial locations, temporal contexts, user's preferences, etc. Most existing studies view the next POI recommendation as a sequence prediction problem while omitting the collaborative signals from other users. Instead, we propose a user-agnostic global trajectory flow map and a novel Graph Enhanced Transformer model (GETNext) to better exploit the extensive collaborative signals for a more accurate next POI prediction, and alleviate the cold start problem in the meantime. GETNext incorporates the global transition patterns, user's general preference, spatio-temporal context, and time-aware category embeddings together into a transformer model to make the prediction of user's future moves. With this design, our model outperforms the state-of-the-art methods with a large margin and also sheds light on the cold start challenges within the spatio-temporal involved recommendation problems.
Leveraging privileged information (PI), or features available during training but not at test time, has recently been shown to be an effective method for addressing label noise. However, the reasons for its effectiveness are not well understood. In this study, we investigate the role played by different properties of the PI in explaining away label noise. Through experiments on multiple datasets with real PI (CIFAR-N/H) and a new large-scale benchmark ImageNet-PI, we find that PI is most helpful when it allows networks to easily distinguish clean from noisy data, while enabling a learning shortcut to memorize the noisy examples. Interestingly, when PI becomes too predictive of the target label, PI methods often perform worse than their no-PI baselines. Based on these findings, we propose several enhancements to the state-of-the-art PI methods and demonstrate the potential of PI as a means of tackling label noise. Finally, we show how we can easily combine the resulting PI approaches with existing no-PI techniques designed to deal with label noise.
Mobile app stores produce a tremendous amount of data in the form of user reviews, which is a huge source of user requirements and sentiments; such reviews allow app developers to proactively address issues in their apps. However, only a small number of reviews capture common issues and sentiments which creates a need for automatically identifying prominent reviews. Unfortunately, most existing work in text ranking and popularity prediction focuses on social contexts where other signals are available, which renders such works ineffective in the context of app reviews. In this work, we propose a new framework, PPrior, that enables proactive prioritization of app issues through identifying prominent reviews (ones predicted to receive a large number of votes in a given time window). Predicting highly-voted reviews is challenging given that, unlike social posts, social network features of users are not available. Moreover, there is an issue of class imbalance, since a large number of user reviews receive little to no votes. PPrior employs a pre-trained T5 model and works in three phases. Phase one adapts the pre-trained T5 model to the user reviews data in a self-supervised fashion. In phase two, we leverage contrastive training to learn a generic and task-independent representation of user reviews. Phase three uses radius neighbors classifier t o m ake t he final predictions. This phase also uses FAISS index for scalability and efficient search. To conduct extensive experiments, we acquired a large dataset of over 2.1 million user reviews from Google Play. Our experimental results demonstrate the effectiveness of the proposed framework when compared against several state-of-the-art approaches. Moreover, the accuracy of PPrior in predicting prominent reviews is comparable to that of experienced app developers.
This work leverages deep learning (DL) techniques in order to do automatic and accurate heart murmur detection from phonocardiogram (PCG) recordings. Two public PCG datasets (CirCor Digiscope 2022 dataset and PCG 2016 dataset) from Physionet online database are utilized to train and test three custom neural networks (NN): a 1D convolutional neural network (CNN), a long short-term memory (LSTM) recurrent neural network (RNN), and a convolutional RNN (C-RNN). Under our proposed method, we first do pre-processing on both datasets in order to prepare the data for the NNs. Key pre-processing steps include the following: denoising, segmentation, re-labeling of noise-only segments, data normalization, and time-frequency analysis of the PCG segments using wavelet scattering transform. To evaluate the performance of the three NNs we have implemented, we conduct four experiments, first three using PCG 2022 dataset, and fourth using PCG 2016 dataset. It turns out that our custom 1D-CNN outperforms other two NNs (LSTM- RNN and C-RNN) as well as the state-of-the-art. Specifically, for experiment E1 (murmur detection using original PCG 2022 dataset), our 1D-CNN model achieves an accuracy of 82.28%, weighted accuracy of 83.81%, F1-score of 65.79%, and and area under receive operating charactertic (AUROC) curve of 90.79%. For experiment E2 (mumur detection using PCG 2022 dataset with unknown class removed), our 1D-CNN model achieves an accuracy of 87.05%, F1-score of 87.72%, and AUROC of 94.4%. For experiment E3 (murmur detection using PCG 2022 dataset with re-labeling of segments), our 1D-CNN model achieves an accuracy of 82.86%, weighted accuracy of 86.30%, F1-score of 81.87%, and AUROC of 93.45%. For experiment E4 (abnormal PCG detection using PCG 2016 dataset), our 1D-CNN model achieves an accuracy of 96.30%, F1-score of 96.29% and AUROC of 98.17%.
Attention mechanism has been widely utilized in speech enhancement (SE) because theoretically it can effectively model the inherent connection of signal both in time domain and spectrum domain. Usually, the span of attention is limited in time domain while the attention in frequency domain spans the whole frequency range. In this paper, we notice that the attention over the whole frequency range hampers the inference for full-band SE and possibly leads to excessive residual noise. To alleviate this problem, we introduce local spectral attention (LSA) into full-band SE model by limiting the span of attention. The ablation test on the state-of-the-art (SOTA) full-band SE model reveals that the local frequency attention can effectively improve overall performance. The improved model achieves the best objective score on the full-band VoiceBank+DEMAND set.