Recognizing a speaker's emotion from their speech can be a key element in emergency call centers. End-to-end deep learning systems for speech emotion recognition now achieve equivalent or even better results than conventional machine learning approaches. In this paper, in order to validate the performance of our neural network architecture for emotion recognition from speech, we first trained and tested it on the widely used corpus accessible by the community, IEMOCAP. We then used the same architecture as the real life corpus, CEMO, composed of 440 dialogs (2h16m) from 485 speakers. The most frequent emotions expressed by callers in these real life emergency dialogues are fear, anger and positive emotions such as relief. In the IEMOCAP general topic conversations, the most frequent emotions are sadness, anger and happiness. Using the same end-to-end deep learning architecture, an Unweighted Accuracy Recall (UA) of 63% is obtained on IEMOCAP and a UA of 45.6% on CEMO, each with 4 classes. Using only 2 classes (Anger, Neutral), the results for CEMO are 76.9% UA compared to 81.1% UA for IEMOCAP. We expect that these encouraging results with CEMO can be improved by combining the audio channel with the linguistic channel. Real-life emotions are clearly more complex than acted ones, mainly due to the large diversity of emotional expressions of speakers. Index Terms-emotion detection, end-to-end deep learning architecture, call center, real-life database, complex emotions.

Accurate segmentation is a crucial step in medical image analysis and applying supervised machine learning to segment the organs or lesions has been substantiated effective. However, it is costly to perform data annotation that provides ground truth labels for training the supervised algorithms, and the high variance of data that comes from different domains tends to severely degrade system performance over cross-site or cross-modality datasets. To mitigate this problem, a novel unsupervised domain adaptation (UDA) method named dispensed Transformer network (DTNet) is introduced in this paper. Our novel DTNet contains three modules. First, a dispensed residual transformer block is designed, which realizes global attention by dispensed interleaving operation and deals with the excessive computational cost and GPU memory usage of the Transformer. Second, a multi-scale consistency regularization is proposed to alleviate the loss of details in the low-resolution output for better feature alignment. Finally, a feature ranking discriminator is introduced to automatically assign different weights to domain-gap features to lessen the feature distribution distance, reducing the performance shift of two domains. The proposed method is evaluated on large fluorescein angiography (FA) retinal nonperfusion (RNP) cross-site dataset with 676 images and a wide used cross-modality dataset from the MM-WHS challenge. Extensive results demonstrate that our proposed network achieves the best performance in comparison with several state-of-the-art techniques.

Background: This paper provides a systematic review of the application of Artificial Intelligence (AI) in the form of Machine Learning (ML) and Deep Learning (DL) techniques in fighting against the effects of novel coronavirus disease (COVID-19). Objective & Methods: The objective is to perform a scoping review on AI for COVID-19 using preferred reporting items of systematic reviews and meta-analysis (PRISMA) guidelines. A literature search was performed for relevant studies published from 1 January 2020 till 27 March 2021. Out of 4050 research papers available in reputed publishers, a full-text review of 440 articles was done based on the keywords of AI, COVID-19, ML, forecasting, DL, X-ray, and Computed Tomography (CT). Finally, 52 articles were included in the result synthesis of this paper. As part of the review, different ML regression methods were reviewed first in predicting the number of confirmed and death cases. Secondly, a comprehensive survey was carried out on the use of ML in classifying COVID-19 patients. Thirdly, different datasets on medical imaging were compared in terms of the number of images, number of positive samples and number of classes in the datasets. The different stages of the diagnosis, including preprocessing, segmentation and feature extraction were also reviewed. Fourthly, the performance results of different research papers were compared to evaluate the effectiveness of DL methods on different datasets. Results: Results show that residual neural network (ResNet-18) and densely connected convolutional network (DenseNet 169) exhibit excellent classification accuracy for X-ray images, while DenseNet-201 has the maximum accuracy in classifying CT scan images. This indicates that ML and DL are useful tools in assisting researchers and medical professionals in predicting, screening and detecting COVID-19.

Convolution neural networks (CNN) are computation intensive to train. It consists of a substantial number of multidimensional dot products between many kernels and inputs. We observe that there are notable similarities among the vectors extracted from inputs (i.e., input vectors). If one input vector is similar to another one, its computations with the kernels are also similar to those of the other and therefore, can be skipped by reusing the already-computed results. Based on this insight, we propose a novel scheme based on locality sensitive hashing (LSH) to exploit the similarity of computations during CNN training in a hardware accelerator. The proposed scheme, called SIMCNN, uses a cache (SIMCACHE) to store LSH signatures of recent input vectors along with the computed results. If the LSH signature of a new input vector matches with that of an already existing vector in the SIMCACHE, the already-computed result is reused for the new vector. SIMCNN is the first work that exploits computational similarity for accelerating CNN training in hardware. The paper presents a detailed design, workflow, and implementation of SIMCNN. Our experimental evaluation with four different deep learning models shows that SIMCNN saves a significant number of computations and therefore, improves training time up to 43%.

Knowledge graphs (KGs) capture knowledge in the form of head--relation--tail triples and are a crucial component in many AI systems. There are two important reasoning tasks on KGs: (1) single-hop knowledge graph completion, which involves predicting individual links in the KG; and (2), multi-hop reasoning, where the goal is to predict which KG entities satisfy a given logical query. Embedding-based methods solve both tasks by first computing an embedding for each entity and relation, then using them to form predictions. However, existing scalable KG embedding frameworks only support single-hop knowledge graph completion and cannot be applied to the more challenging multi-hop reasoning task. Here we present Scalable Multi-hOp REasoning (SMORE), the first general framework for both single-hop and multi-hop reasoning in KGs. Using a single machine SMORE can perform multi-hop reasoning in Freebase KG (86M entities, 338M edges), which is 1,500x larger than previously considered KGs. The key to SMORE's runtime performance is a novel bidirectional rejection sampling that achieves a square root reduction of the complexity of online training data generation. Furthermore, SMORE exploits asynchronous scheduling, overlapping CPU-based data sampling, GPU-based embedding computation, and frequent CPU--GPU IO. SMORE increases throughput (i.e., training speed) over prior multi-hop KG frameworks by 2.2x with minimal GPU memory requirements (2GB for training 400-dim embeddings on 86M-node Freebase) and achieves near linear speed-up with the number of GPUs. Moreover, on the simpler single-hop knowledge graph completion task SMORE achieves comparable or even better runtime performance to state-of-the-art frameworks on both single GPU and multi-GPU settings.

The Transformer architecture has improved the performance of deep learning models in domains such as Computer Vision and Natural Language Processing. Together with better performance come larger model sizes. This imposes challenges to the memory wall of the current accelerator hardware such as GPU. It is never ideal to train large models such as Vision Transformer, BERT, and GPT on a single GPU or a single machine. There is an urgent demand to train models in a distributed environment. However, distributed training, especially model parallelism, often requires domain expertise in computer systems and architecture. It remains a challenge for AI researchers to implement complex distributed training solutions for their models. In this paper, we introduce Colossal-AI, which is a unified parallel training system designed to seamlessly integrate different paradigms of parallelization techniques including data parallelism, pipeline parallelism, multiple tensor parallelism, and sequence parallelism. Colossal-AI aims to support the AI community to write distributed models in the same way as how they write models normally. This allows them to focus on developing the model architecture and separates the concerns of distributed training from the development process. The documentations can be found at https://www.colossalai.org and the source code can be found at https://github.com/hpcaitech/ColossalAI.

Quantifying the pathogenicity of protein variants in human disease-related genes would have a marked effect on clinical decisions, yet the overwhelming majority (over 98%) of these variants still have unknown consequences1–3. In principle, computational methods could support the large-scale interpretation of genetic variants. However, state-of-the-art methods4–10 have relied on training machine learning models on known disease labels. As these labels are sparse, biased and of variable quality, the resulting models have been considered insufficiently reliable11. Here we propose an approach that leverages deep generative models to predict variant pathogenicity without relying on labels. By modelling the distribution of sequence variation across organisms, we implicitly capture constraints on the protein sequences that maintain fitness. Our model EVE (evolutionary model of variant effect) not only outperforms computational approaches that rely on labelled data but also performs on par with, if not better than, predictions from high-throughput experiments, which are increasingly used as evidence for variant classification12–16. We predict the pathogenicity of more than 36 million variants across 3,219 disease genes and provide evidence for the classification of more than 256,000 variants of unknown significance. Our work suggests that models of evolutionary information can provide valuable independent evidence for variant interpretation that will be widely useful in research and clinical settings. A new computational method, EVE, classifies human genetic variants in disease genes using deep generative models trained solely on evolutionary sequences.

This post walks through our submission to the recent Kaggle competition: RSNA-MICCAI Brain Tumor Radiogenomic Classification, which aims at brain tumor detection from 3D MRI scans. I briefly describe the competition and provide data. Later I design a simple training workflow building on several well-established frameworks to produce a robust baseline solution. This baseline received a bronze medal on the private leaderboard even though the public leaderboard score was lower! My takeaway is "simple is better than complex" and it's not wise to overfit the public leaderboards. The RSNA-MICCAI Brain Tumor Radiogenomic Classification competition addresses a fundamental medical screening challenge -- detecting a malignant tumor in the brain.

I recently started an AI-focused educational newsletter, that already has over 100,000 subscribers. TheSequence is a no-BS (meaning no hype, no news etc) ML-oriented newsletter that takes 5 minutes to read. The goal is to keep you up to date with machine learning projects, research papers and concepts. Quantifying trust and fairness is one of the most important challenges to ensure the mainstream adoption of deep learning systems. But what does trust truly means in the context of deep learning systems?

Artificial Intelligence (AI) is beginning to impact science, like physics, by solving some of the most complex, time-consuming, or even impossible problems humans solve. This post discusses some of the applications of artificial intelligence in physics that have been extensively researched. Physicists are also tasked with deciphering deep learning. Deep neural networks are being used in a growing number of applications for automated learning from data, but core theoretical questions regarding how they function remain unanswered. A physics-based solution may assist in closing the gap.