Ivan Marsic

Keyi Li, Sen Yang, Travis M. Sullivan et al.

Process data with confidential information cannot be shared directly in public, which hinders the research in process data mining and analytics. Data encryption methods have been studied to protect the data, but they still may be decrypted, which leads to individual identification. We experimented with different models of representation learning and used the learned model to generate synthetic process data. We introduced an adversarial generative network for process data generation (ProcessGAN) with two Transformer networks for the generator and the discriminator. We evaluated ProcessGAN and traditional models on six real-world datasets, of which two are public and four are collected in medical domains. We used statistical metrics and supervised learning scores to evaluate the synthetic data. We also used process mining to discover workflows for the authentic and synthetic datasets and had medical experts evaluate the clinical applicability of the synthetic workflows. We found that ProcessGAN outperformed traditional sequential models when trained on small authentic datasets of complex processes. ProcessGAN better represented the long-range dependencies between the activities, which is important for complicated processes such as the medical processes. Traditional sequential models performed better when trained on large data of simple processes. We conclude that ProcessGAN can generate a large amount of sharable synthetic process data indistinguishable from authentic data.

Keyi Li, Sen Yang, Travis M. Sullivan et al.

AI-based recommender systems have been successfully applied in many domains (e.g., e-commerce, feeds ranking). Medical experts believe that incorporating such methods into a clinical decision support system may help reduce medical team errors and improve patient outcomes during treatment processes (e.g., trauma resuscitation, surgical processes). Limited research, however, has been done to develop automatic data-driven treatment decision support. We explored the feasibility of building a treatment recommender system to provide runtime next-minute activity predictions. The system uses patient context (e.g., demographics and vital signs) and process context (e.g., activities) to continuously predict activities that will be performed in the next minute. We evaluated our system on a pre-recorded dataset of trauma resuscitation and conducted an ablation study on different model variants. The best model achieved an average F1-score of 0.67 for 61 activity types. We include medical team feedback and discuss the future work.

Chenyang Gao, Yue Gu, Ivan Marsic

In supervised speech separation, permutation invariant training (PIT) is widely used to handle label ambiguity by selecting the best permutation to update the model. Despite its success, previous studies showed that PIT is plagued by excessive label assignment switching in adjacent epochs, impeding the model to learn better label assignments. To address this issue, we propose a novel training strategy, dynamic sample dropout (DSD), which considers previous best label assignments and evaluation metrics to exclude the samples that may negatively impact the learned label assignments during training. Additionally, we include layer-wise optimization (LO) to improve the performance by solving layer-decoupling. Our experiments showed that combining DSD and LO outperforms the baseline and solves excessive label assignment switching and layer-decoupling issues. The proposed DSD and LO approach is easy to implement, requires no extra training sets or steps, and shows generality to various speech separation tasks.

Wenjin Zhang, Keyi Li, Sen Yang et al.

Conventional methods in semi-supervised learning (SSL) often face challenges related to limited data utilization, mainly due to their reliance on threshold-based techniques for selecting high-confidence unlabeled data during training. Various efforts (e.g., FreeMatch) have been made to enhance data utilization by tweaking the thresholds, yet none have managed to use 100% of the available data. To overcome this limitation and improve SSL performance, we introduce \algo, a novel algorithm that fully utilizes unlabeled data to boost semi-supervised learning. \algo integrates a self-supervised learning strategy, i.e., Masked Autoencoder (MAE), that uses all available data to enforce the visual representation learning. This enables the SSL algorithm to leverage all available data, including samples typically filtered out by traditional methods. In addition, we propose a synthetic data training approach to further increase data utilization and improve generalization. These innovations lead \algo to achieve state-of-the-art results on challenging datasets. For instance, on CIFAR-100 with 2 labels per class, STL-10 with 4 labels per class, and Euro-SAT with 2 labels per class, \algo achieves low error rates of 18.71%, 9.47%, and 3.07%, respectively. The code will be made publicly available.

Chenyang Gao, Yue Gu, Ivan Marsic

Speech separation has been studied in time domain because of lower latency and higher performance compared to time-frequency domain. The masking-based method has been mostly used in time domain, and the other common method (mapping-based) has been inadequately studied. We investigate the use of the mapping-based method in the time domain and show that it can perform better on a large training set than the masking-based method. We also investigate the frequent label-switching problem in permutation invariant training (PIT), which results in suboptimal training because the labels selected by PIT differ across training epochs. Our experiment results showed that PIT works well in a shallow separation model, and the label switching occurs for a deeper model. We inferred that layer decoupling may be the reason for the frequent label switching. Therefore, we propose a training strategy based on progressive learning. This approach significantly reduced inconsistent label assignment without added computational complexity or training corpus. By combining this training strategy with the mapping-based method, we significantly improved the separation performance compared to the baseline.

Yanyi Zhang, Xinyu Li, Chunhui Liu et al.

We introduce Video Transformer (VidTr) with separable-attention for video classification. Comparing with commonly used 3D networks, VidTr is able to aggregate spatio-temporal information via stacked attentions and provide better performance with higher efficiency. We first introduce the vanilla video transformer and show that transformer module is able to perform spatio-temporal modeling from raw pixels, but with heavy memory usage. We then present VidTr which reduces the memory cost by 3.3$\times$ while keeping the same performance. To further optimize the model, we propose the standard deviation based topK pooling for attention ($pool_{topK\_std}$), which reduces the computation by dropping non-informative features along temporal dimension. VidTr achieves state-of-the-art performance on five commonly used datasets with lower computational requirement, showing both the efficiency and effectiveness of our design. Finally, error analysis and visualization show that VidTr is especially good at predicting actions that require long-term temporal reasoning.

Jiaojiao Zhao, Yanyi Zhang, Xinyu Li et al.

We propose TubeR: a simple solution for spatio-temporal video action detection. Different from existing methods that depend on either an off-line actor detector or hand-designed actor-positional hypotheses like proposals or anchors, we propose to directly detect an action tubelet in a video by simultaneously performing action localization and recognition from a single representation. TubeR learns a set of tubelet-queries and utilizes a tubelet-attention module to model the dynamic spatio-temporal nature of a video clip, which effectively reinforces the model capacity compared to using actor-positional hypotheses in the spatio-temporal space. For videos containing transitional states or scene changes, we propose a context aware classification head to utilize short-term and long-term context to strengthen action classification, and an action switch regression head for detecting the precise temporal action extent. TubeR directly produces action tubelets with variable lengths and even maintains good results for long video clips. TubeR outperforms the previous state-of-the-art on commonly used action detection datasets AVA, UCF101-24 and JHMDB51-21.

Yanyi Zhang, Xinyu Li, Ivan Marsic

Multi-label activity recognition is designed for recognizing multiple activities that are performed simultaneously or sequentially in each video. Most recent activity recognition networks focus on single-activities, that assume only one activity in each video. These networks extract shared features for all the activities, which are not designed for multi-label activities. We introduce an approach to multi-label activity recognition that extracts independent feature descriptors for each activity and learns activity correlations. This structure can be trained end-to-end and plugged into any existing network structures for video classification. Our method outperformed state-of-the-art approaches on four multi-label activity recognition datasets. To better understand the activity-specific features that the system generated, we visualized these activity-specific features in the Charades dataset.

Jalal Abdulbaqi, Yue Gu, Ivan Marsic

Most current speech enhancement models use spectrogram features that require an expensive transformation and result in phase information loss. Previous work has overcome these issues by using convolutional networks to learn long-range temporal correlations across high-resolution waveforms. These models, however, are limited by memory-intensive dilated convolution and aliasing artifacts from upsampling. We introduce an end-to-end fully-recurrent hourglass-shaped neural network architecture with residual connections for waveform-based single-channel speech enhancement. Our model can efficiently capture long-range temporal dependencies by reducing the features resolution without information loss. Experimental results show that our model outperforms state-of-the-art approaches in six evaluation metrics.

Yanyi Zhang, Xinyu Li, Kaixiang Huang et al.

We present a system for concurrent activity recognition. To extract features associated with different activities, we propose a feature-to-activity attention that maps the extracted global features to sub-features associated with individual activities. To model the temporal associations of individual activities, we propose a transformer-network encoder that models independent temporal associations for each activity. To make the concurrent activity prediction aware of the potential associations between activities, we propose self-attention with an association mask. Our system achieved state-of-the-art or comparable performance on three commonly used concurrent activity detection datasets. Our visualizations demonstrate that our system is able to locate the important spatial-temporal features for final decision making. We also showed that our system can be applied to general multilabel classification problems.

Yue Gu, Kangning Yang, Shiyu Fu et al.

Multimodal affective computing, learning to recognize and interpret human affects and subjective information from multiple data sources, is still challenging because: (i) it is hard to extract informative features to represent human affects from heterogeneous inputs; (ii) current fusion strategies only fuse different modalities at abstract level, ignoring time-dependent interactions between modalities. Addressing such issues, we introduce a hierarchical multimodal architecture with attention and word-level fusion to classify utter-ance-level sentiment and emotion from text and audio data. Our introduced model outperforms the state-of-the-art approaches on published datasets and we demonstrated that our model is able to visualize and interpret the synchronized attention over modalities.

Yue Gu, Shuhong Chen, Ivan Marsic

In this paper, we present a novel deep multimodal framework to predict human emotions based on sentence-level spoken language. Our architecture has two distinctive characteristics. First, it extracts the high-level features from both text and audio via a hybrid deep multimodal structure, which considers the spatial information from text, temporal information from audio, and high-level associations from low-level handcrafted features. Second, we fuse all features by using a three-layer deep neural network to learn the correlations across modalities and train the feature extraction and fusion modules together, allowing optimal global fine-tuning of the entire structure. We evaluated the proposed framework on the IEMOCAP dataset. Our result shows promising performance, achieving 60.4% in weighted accuracy for five emotion categories.

Shuhong Chen, Sen Yang, Moliang Zhou et al.

Adapted from biological sequence alignment, trace alignment is a process mining technique used to visualize and analyze workflow data. Any analysis done with this method, however, is affected by the alignment quality. The best existing trace alignment techniques use progressive guide-trees to heuristically approximate the optimal alignment in O(N2L2) time. These algorithms are heavily dependent on the selected guide-tree metric, often return sum-of-pairs-score-reducing errors that interfere with interpretation, and are computationally intensive for large datasets. To alleviate these issues, we propose process-oriented iterative multiple alignment (PIMA), which contains specialized optimizations to better handle workflow data. We demonstrate that PIMA is a flexible framework capable of achieving better sum-of-pairs score than existing trace alignment algorithms in only O(NL2) time. We applied PIMA to analyzing medical workflow data, showing how iterative alignment can better represent the data and facilitate the extraction of insights from data visualization.

Xinyu Li, Yanyi Zhang, Jianyu Zhang et al.

Process modeling and understanding are fundamental for advanced human-computer interfaces and automation systems. Most recent research has focused on activity recognition, but little has been done on sensor-based detection of process progress. We introduce a real-time, sensor-based system for modeling, recognizing and estimating the progress of a work process. We implemented a multimodal deep learning structure to extract the relevant spatio-temporal features from multiple sensory inputs and used a novel deep regression structure for overall completeness estimation. Using process completeness estimation with a Gaussian mixture model, our system can predict the phase for sequential processes. The performance speed, calculated using completeness estimation, allows online estimation of the remaining time. To train our system, we introduced a novel rectified hyperbolic tangent (rtanh) activation function and conditional loss. Our system was tested on data obtained from the medical process (trauma resuscitation) and sports events (Olympic swimming competition). Our system outperformed the existing trauma-resuscitation phase detectors with a phase detection accuracy of over 86%, an F1-score of 0.67, a completeness estimation error of under 12.6%, and a remaining-time estimation error of less than 7.5 minutes. For the Olympic swimming dataset, our system achieved an accuracy of 88%, an F1-score of 0.58, a completeness estimation error of 6.3% and a remaining-time estimation error of 2.9 minutes.

Xinyu Li, Yanyi Zhang, Ivan Marsic et al.

We introduce a novel, accurate and practical system for real-time people tracking and identification. We used a Kinect V2 sensor for tracking that generates a body skeleton for up to six people in the view. We perform identification using both Kinect and passive RFID, by first measuring the velocity vector of person's skeleton and of their RFID tag using the position of the RFID reader antennas as reference points and then finding the best match between skeletons and tags. We introduce a method for synchronizing Kinect data, which is captured regularly, with irregular or missing RFID data readouts. Our experiments show centimeter-level people tracking resolution with 80% average identification accuracy for up to six people in indoor environments, which meets the needs of many applications. Our system can preserve user privacy and work with different lighting.

Xinyu Li, Yanyi Zhang, Jianyu Zhang et al.

We introduce a system that recognizes concurrent activities from real-world data captured by multiple sensors of different types. The recognition is achieved in two steps. First, we extract spatial and temporal features from the multimodal data. We feed each datatype into a convolutional neural network that extracts spatial features, followed by a long-short term memory network that extracts temporal information in the sensory data. The extracted features are then fused for decision making in the second step. Second, we achieve concurrent activity recognition with a single classifier that encodes a binary output vector in which elements indicate whether the corresponding activity types are currently in progress. We tested our system with three datasets from different domains recorded using different sensors and achieved performance comparable to existing systems designed specifically for those domains. Our system is the first to address the concurrent activity recognition with multisensory data using a single model, which is scalable, simple to train and easy to deploy.