Liming Chen

Sahraoui Dhelim, Tahar Kechadi, Liming Chen et al.

The Metaverse is a virtual environment where users are represented by avatars to navigate a virtual world, which has strong links with the physical one. State-of-the-art Metaverse architectures rely on a cloud-based approach for avatar physics emulation and graphics rendering computation. Such centralized design is unfavorable as it suffers from several drawbacks caused by the long latency required for cloud access, such as low quality visualization. To solve this issue, in this paper, we propose a Fog-Edge hybrid computing architecture for Metaverse applications that leverage an edge-enabled distributed computing paradigm, which makes use of edge devices computing power to fulfil the required computational cost for heavy tasks such as collision detection in virtual universe and computation of 3D physics in virtual simulation. The computational cost related to an entity in the Metaverse such as collision detection or physics emulation are performed at the end-device of the associated physical entity. To prove the effectiveness of the proposed architecture, we simulate a distributed social metaverse application. Simulation results shows that the proposed architecture can reduce the latency by 50% when compared with the legacy cloud-based Metaverse applications.

Jinqiang Wang, Tao Zhu, Liming Chen et al.

Contrastive learning has been applied to Human Activity Recognition (HAR) based on sensor data owing to its ability to achieve performance comparable to supervised learning with a large amount of unlabeled data and a small amount of labeled data. The pre-training task for contrastive learning is generally instance discrimination, which specifies that each instance belongs to a single class, but this will consider the same class of samples as negative examples. Such a pre-training task is not conducive to human activity recognition tasks, which are mainly classification tasks. To address this problem, we follow SimCLR to propose a new contrastive learning framework that negative selection by clustering in HAR, which is called ClusterCLHAR. Compared with SimCLR, it redefines the negative pairs in the contrastive loss function by using unsupervised clustering methods to generate soft labels that mask other samples of the same cluster to avoid regarding them as negative samples. We evaluate ClusterCLHAR on three benchmark datasets, USC-HAD, MotionSense, and UCI-HAR, using mean F1-score as the evaluation metric. The experiment results show that it outperforms all the state-of-the-art methods applied to HAR in self-supervised learning and semi-supervised learning.

Furong Duan, Tao Zhu, Jinqiang Wang et al.

Sensor-based human activity segmentation and recognition are two important and challenging problems in many real-world applications and they have drawn increasing attention from the deep learning community in recent years. Most of the existing deep learning works were designed based on pre-segmented sensor streams and they have treated activity segmentation and recognition as two separate tasks. In practice, performing data stream segmentation is very challenging. We believe that both activity segmentation and recognition may convey unique information which can complement each other to improve the performance of the two tasks. In this paper, we firstly proposes a new multitask deep neural network to solve the two tasks simultaneously. The proposed neural network adopts selective convolution and features multiscale windows to segment activities of long or short time durations. First, multiple windows of different scales are generated to center on each unit of the feature sequence. Then, the model is trained to predict, for each window, the activity class and the offset to the true activity boundaries. Finally, overlapping windows are filtered out by non-maximum suppression, and adjacent windows of the same activity are concatenated to complete the segmentation task. Extensive experiments were conducted on eight popular benchmarking datasets, and the results show that our proposed method outperforms the state-of-the-art methods both for activity recognition and segmentation.

Nyothiri Aung, Tahar Kechadi, Liming Chen et al.

CNNs have been widely applied for medical image analysis. However, limited memory capacity is one of the most common drawbacks of processing high-resolution 3D volumetric data. 3D volumes are usually cropped or downsized first before processing, which can result in a loss of resolution, increase class imbalance, and affect the performance of the segmentation algorithms. In this paper, we propose an end-to-end deep learning approach called IP-UNet. IP-UNet is a UNet-based model that performs multi-class segmentation on Intensity Projection (IP) of 3D volumetric data instead of the memory-consuming 3D volumes. IP-UNet uses limited memory capability for training without losing the original 3D image resolution. We compare the performance of three models in terms of segmentation accuracy and computational cost: 1) Slice-by-slice 2D segmentation of the CT scan images using a conventional 2D UNet model. 2) IP-UNet that operates on data obtained by merging the extracted Maximum Intensity Projection (MIP), Closest Vessel Projection (CVP), and Average Intensity Projection (AvgIP) representations of the source 3D volumes, then applying the UNet model on the output IP images. 3) 3D-UNet model directly reads the 3D volumes constructed from a series of CT scan images and outputs the 3D volume of the predicted segmentation. We test the performance of these methods on 3D volumetric images for automatic breast calcification detection. Experimental results show that IP-Unet can achieve similar segmentation accuracy with 3D-Unet but with much better performance. It reduces the training time by 70\% and memory consumption by 92\%.

Muhammad Aaqib, Aftab Ali, Liming Chen et al.

IoT is one of the fastest-growing technologies and it is estimated that more than a billion devices would be utilized across the globe by the end of 2030. To maximize the capability of these connected entities, trust and reputation among IoT entities is essential. Several trust management models have been proposed in the IoT environment; however, these schemes have not fully addressed the IoT devices features, such as devices role, device type and its dynamic behavior in a smart environment. As a result, traditional trust and reputation models are insufficient to tackle these characteristics and uncertainty risks while connecting nodes to the network. Whilst continuous study has been carried out and various articles suggest promising solutions in constrained environments, research on trust and reputation is still at its infancy. In this paper, we carry out a comprehensive literature review on state-of-the-art research on the trust and reputation of IoT devices and systems. Specifically, we first propose a new structure, namely a new taxonomy, to organize the trust and reputation models based on the ways trust is managed. The proposed taxonomy comprises of traditional trust management-based systems and artificial intelligence-based systems, and combine both the classes which encourage the existing schemes to adapt these emerging concepts. This collaboration between the conventional mathematical and the advanced ML models result in design schemes that are more robust and efficient. Then we drill down to compare and analyse the methods and applications of these systems based on community-accepted performance metrics, e.g. scalability, delay, cooperativeness and efficiency. Finally, built upon the findings of the analysis, we identify and discuss open research issues and challenges, and further speculate and point out future research directions.

Alexandre Chapin, Bruno Machado, Emmanuel Dellandréa et al.

The generalization capabilities of robotic manipulation policies are heavily influenced by the choice of visual representations. Existing approaches typically rely on representations extracted from pre-trained encoders, using two dominant types of features: global features, which summarize an entire image via a single pooled vector, and dense features, which preserve a patch-wise embedding from the final encoder layer. While widely used, both feature types mix task-relevant and irrelevant information, leading to poor generalization under distribution shifts, such as changes in lighting, textures, or the presence of distractors. In this work, we explore an intermediate structured alternative: Slot-Based Object-Centric Representations (SBOCR), which group dense features into a finite set of object-like entities. This representation permits to naturally reduce the noise provided to the robotic manipulation policy while keeping enough information to efficiently perform the task. We benchmark a range of global and dense representations against intermediate slot-based representations, across a suite of simulated and real-world manipulation tasks ranging from simple to complex. We evaluate their generalization under diverse visual conditions, including changes in lighting, texture, and the presence of distractors. Our findings reveal that SBOCR-based policies outperform dense and global representation-based policies in generalization settings, even without task-specific pretraining. These insights suggest that SBOCR is a promising direction for designing visual systems that generalize effectively in dynamic, real-world robotic environments.

Alexandre Chapin, Emmanuel Dellandréa, Liming Chen

Visual foundation models provide strong perceptual features for robotics, but their dense representations lack explicit object-level structure, limiting robustness and contractility in manipulation tasks. We propose STORM (Slot-based Task-aware Object-centric Representation for robotic Manipulation), a lightweight object-centric adaptation module that augments frozen visual foundation models with a small set of semantic-aware slots for robotic manipulation. Rather than retraining large backbones, STORM employs a multi-phase training strategy: object-centric slots are first stabilized through visual--semantic pretraining using language embeddings, then jointly adapted with a downstream manipulation policy. This staged learning prevents degenerate slot formation and preserves semantic consistency while aligning perception with task objectives. Experiments on object discovery benchmarks and simulated manipulation tasks show that STORM improves generalization to visual distractors, and control performance compared to directly using frozen foundation model features or training object-centric representations end-to-end. Our results highlight multi-phase adaptation as an efficient mechanism for transforming generic foundation model features into task-aware object-centric representations for robotic control.

Zezeng Li, Xiaoyu Du, Na Lei et al.

Adversarial attacks exploit the vulnerability of deep models against adversarial samples. Existing point cloud attackers are tailored to specific models, iteratively optimizing perturbations based on gradients in either a white-box or black-box setting. Despite their promising attack performance, they often struggle to produce transferable adversarial samples due to overfitting the specific parameters of surrogate models. To overcome this issue, we shift our focus to the data distribution itself and introduce a novel approach named NoPain, which employs optimal transport (OT) to identify the inherent singular boundaries of the data manifold for cross-network point cloud attacks. Specifically, we first calculate the OT mapping from noise to the target feature space, then identify singular boundaries by locating non-differentiable positions. Finally, we sample along singular boundaries to generate adversarial point clouds. Once the singular boundaries are determined, NoPain can efficiently produce adversarial samples without the need of iterative updates or guidance from the surrogate classifiers. Extensive experiments demonstrate that the proposed end-to-end method outperforms baseline approaches in terms of both transferability and efficiency, while also maintaining notable advantages even against defense strategies. Code and model are available at https://github.com/cognaclee/nopain

Tim Schneider, Guillaume Duret, Cristiana de Farias et al.

Tactile perception has the potential to significantly enhance dexterous robotic manipulation by providing rich local information that can complement or substitute for other sensory modalities such as vision. However, because tactile sensing is inherently local, it is not well-suited for tasks that require broad spatial awareness or global scene understanding on its own. A human-inspired strategy to address this issue is to consider active perception techniques instead. That is, to actively guide sensors toward regions with more informative or significant features and integrate such information over time in order to understand a scene or complete a task. Both active perception and different methods for tactile sensing have received significant attention recently. Yet, despite advancements, both fields lack standardized benchmarks. To bridge this gap, we introduce the Tactile MNIST Benchmark Suite, an open-source, Gymnasium-compatible benchmark specifically designed for active tactile perception tasks, including localization, classification, and volume estimation. Our benchmark suite offers diverse simulation scenarios, from simple toy environments all the way to complex tactile perception tasks using vision-based tactile sensors. Furthermore, we also offer a comprehensive dataset comprising 13,500 synthetic 3D MNIST digit models and 153,600 real-world tactile samples collected from 600 3D printed digits. Using this dataset, we train a CycleGAN for realistic tactile simulation rendering. By providing standardized protocols and reproducible evaluation frameworks, our benchmark suite facilitates systematic progress in the fields of tactile sensing and active perception.

Thomas Duboudin, Emmanuel Dellandréa, Corentin Abgrall et al.

Deep neural networks often fail to generalize outside of their training distribution, in particular when only a single data domain is available during training. While test-time adaptation has yielded encouraging results in this setting, we argue that, to reach further improvements, these approaches should be combined with training procedure modifications aiming to learn a more diverse set of patterns. Indeed, test-time adaptation methods usually have to rely on a limited representation because of the shortcut learning phenomenon: only a subset of the available predictive patterns is learned with standard training. In this paper, we first show that the combined use of existing training-time strategies, and test-time batch normalization, a simple adaptation method, does not always improve upon the test-time adaptation alone on the PACS benchmark. Furthermore, experiments on Office-Home show that very few training-time methods improve upon standard training, with or without test-time batch normalization. We therefore propose a novel approach using a pair of classifiers and a shortcut patterns avoidance loss that mitigates the shortcut learning behavior by reducing the generalization ability of the secondary classifier, using the additional shortcut patterns avoidance loss that encourages the learning of samples specific patterns. The primary classifier is trained normally, resulting in the learning of both the natural and the more complex, less generalizable, features. Our experiments show that our method improves upon the state-of-the-art results on both benchmarks and benefits the most to test-time batch normalization.

Qinkai Yu, Jianyang Xie, Yitian Zhao et al.

Multimodal ophthalmic imaging-based diagnosis integrates color fundus image with optical coherence tomography (OCT) to provide a comprehensive view of ocular pathologies. However, the uneven global distribution of healthcare resources often results in real-world clinical scenarios encountering incomplete multimodal data, which significantly compromises diagnostic accuracy. Existing commonly used pipelines, such as modality imputation and distillation methods, face notable limitations: 1)Imputation methods struggle with accurately reconstructing key lesion features, since OCT lesions are localized, while fundus images vary in style. 2)distillation methods rely heavily on fully paired multimodal training data. To address these challenges, we propose a novel multimodal alignment and fusion framework capable of robustly handling missing modalities in the task of ophthalmic diagnostics. By considering the distinctive feature characteristics of OCT and fundus images, we emphasize the alignment of semantic features within the same category and explicitly learn soft matching between modalities, allowing the missing modality to utilize existing modality information, achieving robust cross-modal feature alignment under the missing modality. Specifically, we leverage the Optimal Transport for multi-scale modality feature alignment: class-wise alignment through predicted class prototypes and feature-wise alignment via cross-modal shared feature transport. Furthermore, we propose an asymmetric fusion strategy that effectively exploits the distinct characteristics of OCT and fundus modalities. Extensive evaluations on three large ophthalmic multimodal datasets demonstrate our model's superior performance under various modality-incomplete scenarios, achieving Sota performance in both complete modality and inter-modality incompleteness conditions. Code is available at https://github.com/Qinkaiyu/RIMA

Shuangjian Li, Tao Zhu, Furong Duan et al.

Wearable sensor-based human activity recognition (HAR) is a critical research domain in activity perception. However, achieving high efficiency and long sequence recognition remains a challenge. Despite the extensive investigation of temporal deep learning models, such as CNNs, RNNs, and transformers, their extensive parameters often pose significant computational and memory constraints, rendering them less suitable for resource-constrained mobile health applications. This study introduces HARMamba, an innovative light-weight and versatile HAR architecture that combines selective bidirectional State Spaces Model and hardware-aware design. To optimize real-time resource consumption in practical scenarios, HARMamba employs linear recursive mechanisms and parameter discretization, allowing it to selectively focus on relevant input sequences while efficiently fusing scan and recompute operations. The model employs independent channels to process sensor data streams, dividing each channel into patches and appending classification tokens to the end of the sequence. It utilizes position embedding to represent the sequence order. The patch sequence is subsequently processed by HARMamba Block, and the classification head finally outputs the activity category. The HARMamba Block serves as the fundamental component of the HARMamba architecture, enabling the effective capture of more discriminative activity sequence features. HARMamba outperforms contemporary state-of-the-art frameworks, delivering comparable or better accuracy with significantly reducing computational and memory demands. It's effectiveness has been extensively validated on 4 publically available datasets namely PAMAP2, WISDM, UNIMIB SHAR and UCI. The F1 scores of HARMamba on the four datasets are 99.74%, 99.20%, 88.23% and 97.01%, respectively.

Mohanad Odema, Luke Chen, Hyoukjun Kwon et al.

Emerging multi-model workloads with heavy models like recent large language models significantly increased the compute and memory demands on hardware. To address such increasing demands, designing a scalable hardware architecture became a key problem. Among recent solutions, the 2.5D silicon interposer multi-chip module (MCM)-based AI accelerator has been actively explored as a promising scalable solution due to their significant benefits in the low engineering cost and composability. However, previous MCM accelerators are based on homogeneous architectures with fixed dataflow, which encounter major challenges from highly heterogeneous multi-model workloads due to their limited workload adaptivity. Therefore, in this work, we explore the opportunity in the heterogeneous dataflow MCM AI accelerators. We identify the scheduling of multi-model workload on heterogeneous dataflow MCM AI accelerator is an important and challenging problem due to its significance and scale, which reaches O(10^56) even for a two-model workload on 6x6 chiplets. We develop a set of heuristics to navigate the huge scheduling space and codify them into a scheduler, SCAR, with advanced techniques such as inter-chiplet pipelining. Our evaluation on ten multi-model workload scenarios for datacenter multitenancy and AR/VR use-cases has shown the efficacy of our approach, achieving on average 27.6% and 29.6% less energy-delay product (EDP) for the respective applications settings compared to homogeneous baselines.

Jiyu Tian, Mingchu Li, Liming Chen et al.

Anomaly detection for cyber-physical systems (ADCPS) is crucial in identifying faults and potential attacks by analyzing the time series of sensor measurements and actuator states. However, current methods lack adaptation to data distribution shifts in both temporal and spatial dimensions as cyber-physical systems evolve. To tackle this issue, we propose an incremental meta-learning-based approach, namely iADCPS, which can continuously update the model through limited evolving normal samples to reconcile the distribution gap between evolving and historical time series. Specifically, We first introduce a temporal mixup strategy to align data for data-level generalization which is then combined with the one-class meta-learning approach for model-level generalization. Furthermore, we develop a non-parametric dynamic threshold to adaptively adjust the threshold based on the probability density of the abnormal scores without any anomaly supervision. We empirically evaluate the effectiveness of the iADCPS using three publicly available datasets PUMP, SWaT, and WADI. The experimental results demonstrate that our method achieves 99.0%, 93.1%, and 78.7% F1-Score, respectively, which outperforms the state-of-the-art (SOTA) ADCPS method, especially in the context of the evolving CPSs.

Lingkun Luo, Shiqiang Hu, Liming Chen

In domain adaptation (DA), the effectiveness of deep learning-based models is often constrained by batch learning strategies that fail to fully apprehend the global statistical and geometric characteristics of data distributions. Addressing this gap, we introduce 'Global Awareness Enhanced Domain Adaptation' (GAN-DA), a novel approach that transcends traditional batch-based limitations. GAN-DA integrates a unique predefined feature representation (PFR) to facilitate the alignment of cross-domain distributions, thereby achieving a comprehensive global statistical awareness. This representation is innovatively expanded to encompass orthogonal and common feature aspects, which enhances the unification of global manifold structures and refines decision boundaries for more effective DA. Our extensive experiments, encompassing 27 diverse cross-domain image classification tasks, demonstrate GAN-DA's remarkable superiority, outperforming 24 established DA methods by a significant margin. Furthermore, our in-depth analyses shed light on the decision-making processes, revealing insights into the adaptability and efficiency of GAN-DA. This approach not only addresses the limitations of existing DA methodologies but also sets a new benchmark in the realm of domain adaptation, offering broad implications for future research and applications in this field.

Xiaoyan Kui, Haonan Yan, Qinsong Li et al.

Graph neural networks have achieved remarkable success in learning graph representations, especially graph Transformer, which has recently shown superior performance on various graph mining tasks. However, graph Transformer generally treats nodes as tokens, which results in quadratic complexity regarding the number of nodes during self-attention computation. The graph MLP Mixer addresses this challenge by using the efficient MLP Mixer technique from computer vision. However, the time-consuming process of extracting graph tokens limits its performance. In this paper, we present a novel architecture named ChebMixer, a newly graph MLP Mixer that uses fast Chebyshev polynomials-based spectral filtering to extract a sequence of tokens. Firstly, we produce multiscale representations of graph nodes via fast Chebyshev polynomial-based spectral filtering. Next, we consider each node's multiscale representations as a sequence of tokens and refine the node representation with an effective MLP Mixer. Finally, we aggregate the multiscale representations of nodes through Chebyshev interpolation. Owing to the powerful representation capabilities and fast computational properties of MLP Mixer, we can quickly extract more informative node representations to improve the performance of downstream tasks. The experimental results prove our significant improvements in a variety of scenarios ranging from graph node classification to medical image segmentation.

Lingkun Luo, Shiqiang Hu, Liming Chen

Deep learning-based domain adaptation (DA) methods have shown strong performance by learning transferable representations. However, their reliance on mini-batch training limits global distribution modeling, leading to unstable alignment and suboptimal generalization. We propose Global Variational Inference Enhanced Domain Adaptation (GVI-DA), a framework that learns continuous, class-conditional global priors via variational inference to enable structure-aware cross-domain alignment. GVI-DA minimizes domain gaps through latent feature reconstruction, and mitigates posterior collapse using global codebook learning with randomized sampling. It further improves robustness by discarding low-confidence pseudo-labels and generating reliable target-domain samples. Extensive experiments on four benchmarks and thirty-eight DA tasks demonstrate consistent state-of-the-art performance. We also derive the model's evidence lower bound (ELBO) and analyze the effects of prior continuity, codebook size, and pseudo-label noise tolerance. In addition, we compare GVI-DA with diffusion-based generative frameworks in terms of optimization principles and efficiency, highlighting both its theoretical soundness and practical advantages.

Qi Qiu, Tao Zhu, Furong Duan et al.

Inertial Measurement Unit (IMU) sensors are widely employed for Human Activity Recognition (HAR) due to their portability, energy efficiency, and growing research interest. However, a significant challenge for IMU-HAR models is achieving robust generalization performance across diverse users. This limitation stems from substantial variations in data distribution among individual users. One primary reason for this distribution disparity lies in the representation of IMU sensor data in the local coordinate system, which is susceptible to subtle user variations during IMU wearing. To address this issue, we propose a novel approach that extracts a global view representation based on the characteristics of IMU data, effectively alleviating the data distribution discrepancies induced by wearing styles. To validate the efficacy of the global view representation, we fed both global and local view data into model for experiments. The results demonstrate that global view data significantly outperforms local view data in cross-user experiments. Furthermore, we propose a Multi-view Supervised Network (MVFNet) based on Shuffling to effectively fuse local view and global view data. It supervises the feature extraction of each view through view division and view shuffling, so as to avoid the model ignoring important features as much as possible. Extensive experiments conducted on OPPORTUNITY and PAMAP2 datasets demonstrate that the proposed algorithm outperforms the current state-of-the-art methods in cross-user HAR.

Wenyong Han, Tao Zhu Member, Liming Chen et al.

The massive generation of time-series data by largescale Internet of Things (IoT) devices necessitates the exploration of more effective models for multivariate time-series forecasting. In previous models, there was a predominant use of the Channel Dependence (CD) strategy (where each channel represents a univariate sequence). Current state-of-the-art (SOTA) models primarily rely on the Channel Independence (CI) strategy. The CI strategy treats all channels as a single channel, expanding the dataset to improve generalization performance and avoiding inter-channel correlation that disrupts long-term features. However, the CI strategy faces the challenge of interchannel correlation forgetting. To address this issue, we propose an innovative Mixed Channels strategy, combining the data expansion advantages of the CI strategy with the ability to counteract inter-channel correlation forgetting. Based on this strategy, we introduce MCformer, a multivariate time-series forecasting model with mixed channel features. The model blends a specific number of channels, leveraging an attention mechanism to effectively capture inter-channel correlation information when modeling long-term features. Experimental results demonstrate that the Mixed Channels strategy outperforms pure CI strategy in multivariate time-series forecasting tasks.

Shuangjian Li, Tao Zhu, Mingxing Nie et al.

Traditional deep learning methods struggle to simultaneously segment, recognize, and forecast human activities from sensor data. This limits their usefulness in many fields such as healthcare and assisted living, where real-time understanding of ongoing and upcoming activities is crucial. This paper introduces P2LHAP, a novel Patch-to-Label Seq2Seq framework that tackles all three tasks in a efficient single-task model. P2LHAP divides sensor data streams into a sequence of "patches", served as input tokens, and outputs a sequence of patch-level activity labels including the predicted future activities. A unique smoothing technique based on surrounding patch labels, is proposed to identify activity boundaries accurately. Additionally, P2LHAP learns patch-level representation by sensor signal channel-independent Transformer encoders and decoders. All channels share embedding and Transformer weights across all sequences. Evaluated on three public datasets, P2LHAP significantly outperforms the state-of-the-art in all three tasks, demonstrating its effectiveness and potential for real-world applications.

Jiangtao Wang, Bin Guo, Liming Chen

Though technical advance of artificial intelligence and machine learning has enabled many promising intelligent systems, many computing tasks are still not able to be fully accomplished by machine intelligence. Motivated by the complementary nature of human and machine intelligence, an emerging trend is to involve humans in the loop of machine learning and decision-making. In this paper, we provide a macro-micro review of human-in-the-loop machine learning. We first describe major machine learning challenges which can be addressed by human intervention in the loop. Then we examine closely the latest research and findings of introducing humans into each step of the lifecycle of machine learning. Finally, we analyze current research gaps and point out future research directions.

Sahraoui Dhelim, Liming Chen, Huansheng Ning et al.

Death by suicide is the seventh leading death cause worldwide. The recent advancement in Artificial Intelligence (AI), specifically AI applications in image and voice processing, has created a promising opportunity to revolutionize suicide risk assessment. Subsequently, we have witnessed fast-growing literature of research that applies AI to extract audiovisual non-verbal cues for mental illness assessment. However, the majority of the recent works focus on depression, despite the evident difference between depression symptoms and suicidal behavior and non-verbal cues. This paper reviews recent works that study suicide ideation and suicide behavior detection through audiovisual feature analysis, mainly suicidal voice/speech acoustic features analysis and suicidal visual cues. Automatic suicide assessment is a promising research direction that is still in the early stages. Accordingly, there is a lack of large datasets that can be used to train machine learning and deep learning models proven to be effective in other, similar tasks.

Jinqiang Wang, Tao Zhu, Jingyuan Gan et al.

While deep learning has contributed to the advancement of sensor-based Human Activity Recognition (HAR), it is usually a costly and challenging supervised task with the needs of a large amount of labeled data. To alleviate this issue, contrastive learning has been applied for sensor-based HAR. Data augmentation is an essential part of contrastive learning and has a significant impact on the performance of downstream tasks. However, current popular augmentation methods do not achieve competitive performance in contrastive learning for sensor-based HAR. Motivated by this issue, we propose a new sensor data augmentation method by resampling, which simulates more realistic activity data by varying the sampling frequency to maximize the coverage of the sampling space. In addition, we extend MoCo, a popular contrastive learning framework, to MoCoHAR for HAR. The resampling augmentation method will be evaluated on two contrastive learning frameworks, SimCLRHAR and MoCoHAR, using UCI-HAR, MotionSensor, and USC-HAD datasets. The experiment results show that the resampling augmentation method outperforms all state-of-the-art methods under a small amount of labeled data, on SimCLRHAR and MoCoHAR, with mean F1-score as the evaluation metric. The results also demonstrate that not all data augmentation methods have positive effects in the contrastive learning framework.

Sahraoui Dhelim, Liming Luke Chen, Nyothiri Aung et al.

Personality-aware recommendation systems have been proven to achieve high accuracy compared to conventional recommendation systems. In addition to that, personality-aware recommendation systems could help alleviate cold start and data sparsity problems. Most of the existing works use Big-Five personality model to represent the user's personality, this is due to the popularity of Big-Five model in the literature of psychology. However, from personality computing perspective, the choice of the most suitable personality model that satisfy the requirements of the recommendation application and the recommended content type still needs further investigation. In this paper, we study and compare four personality-aware recommendation systems based on different personality models, namely Big-Five, Eysenck and HEXACO from the personality traits theory, and Myers-Briggs Type Indicator (MPTI) from the personality types theory. Following that, we propose a hybrid personality model for recommendation that takes advantage of the personality traits models, as well as the personality types models. Through extensive experiments on recommendation dataset, we prove the efficiency of the proposed model, especially in cold start settings.

Sahraoui Dhelim, Liming Luke Chen, Huansheng Ning et al.

Online social media provides a channel for monitoring people's social behaviors and their mental distress. Due to the restrictions imposed by COVID-19 people are increasingly using online social networks to express their feelings. Consequently, there is a significant amount of diverse user-generated social media content. However, COVID-19 pandemic has changed the way we live, study, socialize and recreate and this has affected our well-being and mental health problems. There are growing researches that leverage online social media analysis to detect and assess user's mental status. In this paper, we survey the literature of social media analysis for mental disorders detection, with a special focus on the studies conducted in the context of COVID-19 during 2020-2021. Firstly, we classify the surveyed studies in terms of feature extraction types, varying from language usage patterns to aesthetic preferences and online behaviors. Secondly, we explore detection methods used for mental disorders detection including machine learning and deep learning detection methods. Finally, we discuss the challenges of mental disorder detection using social media data, including the privacy and ethical concerns, as well as the technical challenges of scaling and deploying such systems at large scales, and discuss the learnt lessons over the last few years.

Dimitri Gominski, Valérie Gouet-Brunet, Liming Chen

With impressive results in applications relying on feature learning, deep learning has also blurred the line between algorithm and data. Pick a training dataset, pick a backbone network for feature extraction, and voilà ; this usually works for a variety of use cases. But the underlying hypothesis that there exists a training dataset matching the use case is not always met. Moreover, the demand for interconnections regardless of the variations of the content calls for increasing generalization and robustness in features. An interesting application characterized by these problematics is the connection of historical and cultural databases of images. Through the seemingly simple task of instance retrieval, we propose to show that it is not trivial to pick features responding well to a panel of variations and semantic content. Introducing a new enhanced version of the Alegoria benchmark, we compare descriptors using the detailed annotations. We further give insights about the core problems in instance retrieval, testing four state-of-the-art additional techniques to increase performance.

Dimitri Gominski, Valérie Gouet-Brunet, Liming Chen

Advances in high resolution remote sensing image analysis are currently hampered by the difficulty of gathering enough annotated data for training deep learning methods, giving rise to a variety of small datasets and associated dataset-specific methods. Moreover, typical tasks such as classification and retrieval lack a systematic evaluation on standard benchmarks and training datasets, which make it hard to identify durable and generalizable scientific contributions. We aim at unifying remote sensing image retrieval and classification with a new large-scale training and testing dataset, SF300, including both vertical and oblique aerial images and made available to the research community, and an associated fine-tuning method. We additionally propose a new adversarial fine-tuning method for global descriptors. We show that our framework systematically achieves a boost of retrieval and classification performance on nine different datasets compared to an ImageNet pretrained baseline, with currently no other method to compare to.

Kexin Lv, Fan He, Xiaolin Huang et al.

Nowadays, more and more datasets are stored in a distributed way for the sake of memory storage or data privacy. The generalized eigenvalue problem (GEP) plays a vital role in a large family of high-dimensional statistical models. However, the existing distributed method for eigenvalue decomposition cannot be applied in GEP for the divergence of the empirical covariance matrix. Here we propose a general distributed GEP framework with one-shot communication for GEP. If the symmetric data covariance has repeated eigenvalues, e.g., in canonical component analysis, we further modify the method for better convergence. The theoretical analysis on approximation error is conducted and the relation to the divergence of the data covariance, the eigenvalues of the empirical data covariance, and the number of local servers is analyzed. Numerical experiments also show the effectiveness of the proposed algorithms.

Mathilde Guillemot, Catherine Heusele, Rodolphe Korichi et al.

The lack of transparency of neural networks stays a major break for their use. The Layerwise Relevance Propagation technique builds heat-maps representing the relevance of each input in the model s decision. The relevance spreads backward from the last to the first layer of the Deep Neural Network. Layer-wise Relevance Propagation does not manage normalization layers, in this work we suggest a method to include normalization layers. Specifically, we build an equivalent network fusing normalization layers and convolutional or fully connected layers. Heatmaps obtained with our method on MNIST and CIFAR 10 datasets are more accurate for convolutional layers. Our study also prevents from using Layerwise Relevance Propagation with networks including a combination of connected layers and normalization layer.

Thomas Duboudin, Maxime Petit, Liming Chen

We propose a procedural fruit tree rendering framework, based on Blender and Python scripts allowing to generate quickly labeled dataset (i.e. including ground truth semantic segmentation). It is designed to train image analysis deep learning methods (e.g. in a robotic fruit harvesting context), where real labeled training datasets are usually scarce and existing synthetic ones are too specialized. Moreover, the framework includes the possibility to introduce parametrized variations in the model (e.g. lightning conditions, background), producing a dataset with embedded Domain Randomization aspect.

Erinc Merdivan, Anastasios Vafeiadis, Dimitrios Kalatzis et al.

We propose a new approach to natural language understanding in which we consider the input text as an image and apply 2D Convolutional Neural Networks to learn the local and global semantics of the sentences from the variations ofthe visual patterns of words. Our approach demonstrates that it is possible to get semantically meaningful features from images with text without using optical character recognition and sequential processing pipelines, techniques that traditional Natural Language Understanding algorithms require. To validate our approach, we present results for two applications: text classification and dialog modeling. Using a 2D Convolutional Neural Network, we were able to outperform the state-of-art accuracy results of non-Latin alphabet-based text classification and achieved promising results for eight text classification datasets. Furthermore, our approach outperformed the memory networks when using out of vocabulary entities fromtask 4 of the bAbI dialog dataset.

Asim Jan, Huaxiong Ding, Hongying Meng et al.

Meaningful facial parts can convey key cues for both facial action unit detection and expression prediction. Textured 3D face scan can provide both detailed 3D geometric shape and 2D texture appearance cues of the face which are beneficial for Facial Expression Recognition (FER). However, accurate facial parts extraction as well as their fusion are challenging tasks. In this paper, a novel system for 3D FER is designed based on accurate facial parts extraction and deep feature fusion of facial parts. In particular, each textured 3D face scan is firstly represented as a 2D texture map and a depth map with one-to-one dense correspondence. Then, the facial parts of both texture map and depth map are extracted using a novel 4-stage process consists of facial landmark localization, facial rotation correction, facial resizing, facial parts bounding box extraction and post-processing procedures. Finally, deep fusion Convolutional Neural Networks (CNNs) features of all facial parts are learned from both texture maps and depth maps, respectively and nonlinear SVMs are used for expression prediction. Experiments are conducted on the BU-3DFE database, demonstrating the effectiveness of combing different facial parts, texture and depth cues and reporting the state-of-the-art results in comparison with all existing methods under the same setting.

Ismini Psychoula, Erinc Merdivan, Deepika Singh et al.

In the era of Internet of Things (IoT) technologies the potential for privacy invasion is becoming a major concern especially in regards to healthcare data and Ambient Assisted Living (AAL) environments. Systems that offer AAL technologies make extensive use of personal data in order to provide services that are context-aware and personalized. This makes privacy preservation a very important issue especially since the users are not always aware of the privacy risks they could face. A lot of progress has been made in the deep learning field, however, there has been lack of research on privacy preservation of sensitive personal data with the use of deep learning. In this paper we focus on a Long Short Term Memory (LSTM) Encoder-Decoder, which is a principal component of deep learning, and propose a new encoding technique that allows the creation of different AAL data views, depending on the access level of the end user and the information they require access to. The efficiency and effectiveness of the proposed method are demonstrated with experiments on a simulated AAL dataset. Qualitatively, we show that the proposed model learns privacy operations such as disclosure, deletion and generalization and can perform encoding and decoding of the data with almost perfect recovery.

Matthieu Grard, Romain Brégier, Florian Sella et al.

With more and more household objects built on planned obsolescence and consumed by a fast-growing population, hazardous waste recycling has become a critical challenge. Given the large variability of household waste, current recycling platforms mostly rely on human operators to analyze the scene, typically composed of many object instances piled up in bulk. Helping them by robotizing the unitary extraction is a key challenge to speed up this tedious process. Whereas supervised deep learning has proven very efficient for such object-level scene understanding, e.g., generic object detection and segmentation in everyday scenes, it however requires large sets of per-pixel labeled images, that are hardly available for numerous application contexts, including industrial robotics. We thus propose a step towards a practical interactive application for generating an object-oriented robotic grasp, requiring as inputs only one depth map of the scene and one user click on the next object to extract. More precisely, we address in this paper the middle issue of object seg-mentation in top views of piles of bulk objects given a pixel location, namely seed, provided interactively by a human operator. We propose a twofold framework for generating edge-driven instance segments. First, we repurpose a state-of-the-art fully convolutional object contour detector for seed-based instance segmentation by introducing the notion of edge-mask duality with a novel patch-free and contour-oriented loss function. Second, we train one model using only synthetic scenes, instead of manually labeled training data. Our experimental results show that considering edge-mask duality for training an encoder-decoder network, as we suggest, outperforms a state-of-the-art patch-based network in the present application context.