LIREM: A Generic Framework for Effective Online Video Novelty Detection

Novelty detection is a very desirable additional feature of any practical classification or forecasting system. Novelty and rare patterns detection is the main objective in such applications as fault/abnormality discovery in complex technical and biological systems, fraud detection and risk management in financial and insurance industry. Although many interdisciplinary approaches for rare event modeling and novelty detection have been proposed, significant data incompleteness due to the nature of the problem makes it difficult to find a universal solution. Even more challenging and much less formalized problem is novelty detection in complex strategies and models where practical performance criteria are usually multi‐objective and the best state‐of‐the‐art solution is often not known due to the complexity of the task and/or proprietary nature of the application area. For example, it is much more difficult to detect a series of small insider trading or other illegal transactions mixed with valid operations...

Supervised machine learning is a powerful and widely used method for analyzing high-content screening data. Despite its accuracy, efficiency, and versatility, supervised machine learning has drawbacks, most notably its dependence on a priori knowledge of expected phenotypes and time-consuming classifier training. We provide a solution to these limitations with CellCognition Explorer, a generic novelty detection and deep learning framework. Application to several large-scale screening data sets on nuclear and mitotic cell morphologies demonstrates that CellCognition Explorer enables discovery of rare phenotypes without user training, which has broad implications for improved assay development in high-content screening.

The increasing energy consumption of heating, ventilation and air conditioning (HVAC) systems is one of the main concerns in the building sector. Fault detection technologies are now indispensable for energy efficiency and performance improvement. In this paper, a methodology for the robust and reliable fault detection and diagnosis is presented as a two-stage framework composed by an offline stage where the models are built and an online stage that is constantly receiving new samples. The system includes a novelty detection scheme developed using one-class support vector machines (OC-SVM) and a classifier built using SVM. The proposed strategy is applied to a dataset for a single-zone constant air volume air handling unit. The experimental results show that the novelty detection stage adds robustness layer to the typical classification scheme.

This study aims to evaluate the informational quality of oral cancer-related videos on YouTube and Bilibili. A total of 300 oral videos that met the inclusion and exclusion criteria were selected for analysis. The selection comprised 150 videos from 111 uploaders on YouTube and 150 videos from 134 uploaders on Bilibili. YouTube videos received a greater number of views and likes, while there was no significant difference in average likes per 30 days or comments between the two platforms. The majority of YouTube uploaders were hospitals/non-profit organizations (66.7%) and for-profit companies (17.1%), while Bilibili uploaders were mainly self-media (55.2%) and doctors (29.1%). YouTube videos covered a broader range of topics compared to Bilibili videos. Though solo narration was the most prevalent video style across both platforms, YouTube exhibited a higher preference for TV shows/documentaries (31.3%). Video quality was assessed using the four tools: Though mDISCERN (modified DISCERN) and PEMAT-Actionability (Patient Education Materials Assessment Tool) were similar across platforms, YouTube videos scored higher on PEMAT-Understandability, VIQI (Video Information and Quality Index), and GQS (Global Quality Score) in comparison to Bilibili videos. Videos produced by health professionals were considered more reliable. Spearman correlation analysis revealed no strong relationships between video quality and audience interaction. In conclusion, YouTube videos exhibited higher audience engagement and video quality, yet improvements are needed on both platforms. In order to promote high-quality health information, it is essential to encourage the development of more professional content creators and to optimize platform algorithms.

Spatio-temporal information is crucial to human action recognition. Inspired by the primary visual cortex of the human brain, a new end-to-end action recognition method based on deep learning was proposed to extract pure spatiotemporal information in videos. It is named moving human focus inference model that can capture long-term temporal information with low time costs and eliminates the interference of dynamic background. It obtains the temporal and spatial information in the video from the temporal pathway with the focus block and spatial pathway, respectively. Then the temporal and spatial information are merged as spatio-temporal information by fusion region. Finally, the softmax layer is used to recognize the actions, followed by the bayes inference block that is used to correct the recognition results through the methods of maximum posterior probability inference. The model is evaluated on two challenge datasets: UCF101 and HMDB51. The experimental results show that the approach has better accuracy and generalization ability compared to the state-of-the-art methods.

Recent advances in "track-anything" models have significantly improved fine-grained video understanding by simultaneously handling multiple video segmentation and tracking tasks. However, existing models often struggle with robust and efficient temporal propagation. To address these challenges, we propose the Sparse Spatio-Temporal Propagation (SSTP) method, which achieves robust and efficient unified video segmentation by selectively leveraging key spatio-temporal features in videos. Specifically, we design a dynamic 3D spatio-temporal convolution to aggregate global multi-frame spatio-temporal information into memory frames during memory construction. Additionally, we introduce a spatio-temporal aggregation reading strategy to efficiently aggregate the relevant spatio-temporal features from multiple memory frames during memory retrieval. By combining SSTP with an image segmentation foundation model, such as the segment anything model, our method effectively addresses multiple data-scarce video segmentation tasks. Our experimental results demonstrate state-of-the-art performance on five video segmentation tasks across eleven datasets, outperforming both task-specific and unified methods. Notably, SSTP exhibits strong robustness in handling sparse, low-frame-rate videos, making it well-suited for real-world applications.

Convolutional Neural Networks(CNNs) have achieved great success for object recognition in still images. However, CNNs can't make evident improvement for action recognition in videos, one reason is that many current network architectures are relatively shallow compared with deep models in image domain, and the other reason is that CNNs can't capture effective long-term motion information from videos. Encouraged by the good performance of Residual Network-s(ResNets) for training extremely deep models, and Long-term Recurrent Convolutional Networks(LSTMs) for dealing with tasks involving sequences, we presented an action recognition method based on a two-stream architecture, with 2D ResNets with LSTMs in one stream and designed 3D ResNets with LSTMs in the other stream, which can combine appearance and motion information better. Especially, our proposed method first learns spatiotemporal features of videos through the Residual networks, then models complex temporal dynamics by the Long-term Recurrent Convolutional networks, and with a softmax layer on the top of two streams, the final classification results can be predicted by fusing scores of each stream with weights on score distribution. Furthermore, for better reducing the influence of redundant background information in videos for recognition results, we also applied a center extraction method to generate central regions of videos instead of an entire video into a visual representation. On two video action benchmarks of UCF101 and HMDB51, our method achieved promising performance compared with state-of-the-art.

Well-performed Video quality assessment (VQA) method should be consistent with human visual systems for better prediction accuracy. In this paper, we propose a VQA method using motion-compensated temporal filtering (MCTF) and manifold feature similarity. To be more specific, a group of frames (GoF) is first decomposed into a temporal high-pass component (HPC) and a temporal low-pass component (LPC) by MCTF. Following this, manifold feature learning (MFL) and phase congruency (PC) are used to predict the quality of temporal LPC and temporal HPC respectively. The quality measures of the LPC and the HPC are then combined as GoF quality. A temporal pooling strategy is subsequently used to integrate GoF qualities into an overall video quality. The proposed VQA method appropriately processes temporal information in video by MCTF and temporal pooling strategy, and simulate human visual perception by MFL. Experiments on publicly available video quality database showed that in comparison with several state-of-the-art VQA methods, the proposed VQA method achieves better consistency with subjective video quality and can predict video quality more accurately.

ObjectiveTo conduct a multidimensional evaluation of endometrial cancer (EC)-related videos on major Chinese short-video platforms.MethodsA cross-sectional study conducted on May 8, 2025 analyzed 226 eligible EC-related videos from TikTok, Rednote, and Bilibili. Video quality was assessed using Global Quality Scale (GQS) and Video Information and Quality Index (VIQI), reliability using modified DISCERN (mDISCERN), and understandability and actionability using Patient Education Materials Assessment Tool (PEMAT).ResultsAmong 226 videos, TikTok emphasized symptoms/risk factors and scored highest in engagement, reliability (mDISCERN=2.0, P=0.002), and understandability (88%). Bilibili led in VIQI (median=17.0) but had the lowest understandability (67%). Professional videos outperformed patient-generated content (all P<0.05). Video length correlated positively with quality but negatively with engagement and understandability (all P<0.01).ConclusionsEC-related videos vary widely in quality and lack consistent actionability. Strategic content design and platform-level verification are needed to improve reliability and public health impact.

This study aimed to evaluate the content and quality of YouTube videos of zygomatic implants. According to Google Trends (2021), "zygomatic implant" was the most preferred keyword related to the topic. Therefore, in this study "zygomatic implant" was used as a keyword for the video search. Demographic characteristics such as the number of views, likes/dislikes, comments, video duration, number of days after upload, uploaders, and target audiences of the videos were evaluated. To evaluate the accuracy and content quality of videos (available from YouTube), the video information and quality index (VIQI) and global quality scale (GQS) were used. Statistical analyses were performed using the Kruskal-Wallis test, Mann-Whitney U test, chi-square test, Fisher's exact chi-square test, Yates continuity correction, and Spearman correlation analysis (p < 0.05). A total of 151 videos were searched; 90 met all inclusion criteria. According to the video content score, 78.9% of the videos were identified as low content, 20% as moderate, and 1.1% as high content. There was no statistical difference between the groups in video demographic characteristics (p> 0.001). Conversely, information flow, accuracy of information, video quality and precision, and total VIQI scores were statistically different between the groups. The moderate-content group had a higher GQS score than the low-content group (p < 0.001). The videos were mainly uploaded (40%) from hospitals and universities. Most videos were targeted toward professionals (46.75%). Low-content videos had higher ratings than the moderate- and high-content videos. Most YouTube videos on zygomatic implants showed low-content quality. This implies that YouTube is not a reliable source of information on zygomatic implants. Dentists, prosthodontists, and oral and maxillofacial surgeons should be aware of the content of video-sharing platforms and take responsibility for enriching video content.

In multimedia-based e-learning systems, the accessibility and searchability of most lecture video content is still insufficient due to the unscripted and spontaneous speech of the speakers. Thus, it is very desirable to enable people to navigate and access specific topics within lecture videos by performing an automatic topic-wise video segmentation. This problem becomes even more challenging when the quality of such lecture videos is not sufficiently high. To this end, we first present the ATLAS system that has two main novelties: (i) a SVM hmm model is proposed to learn temporal transition cues and (ii) a fusion scheme is suggested to combine transition cues extracted from heterogeneous information of lecture videos. Subsequently, considering that contextual information is very useful in determining knowledge structures, we present the TRACE system to automatically perform such a segmentation based on a linguistic approach using Wikipedia texts. TRACE has two main contributions: (i) the extraction of a novel linguistic-based Wikipedia feature to segment lecture videos efficiently, and (ii) the investigation of the late fusion of video segmentation results derived from state-of-the-art algorithms. Specifically for the late fusion, we combine confidence scores produced by the models constructed from visual, transcriptional, and Wikipedia features. According to our experiments on lecture videos from VideoLectures.NET and NPTEL, the proposed algorithms in the ATLAS and TRACE systems segment knowledge structures more accurately compared to existing state-of-the-art algorithms.

It is substantiated that important in this direction is the issue of ensuring security and the necessary quality of information service provision, including the timely transfer of various types of information with sufficient quality and reliability. It is established that video information is of special importance here. However, the required bit rate level of video stream transmission is possible only in conditions of low quality of video segments. Therefore, a significant imbalance between the information load and the bandwidth of information communication networks contributes. This becomes a causal factor in the loss of the level of credibility and adequacy of video information. So, it is shown that the current issue of research concerns the improvement of the quality of providing video information services using the mobile segment on the basis of eliminating the imbalance between the significant intensity of video information flows and the limited productivity of information communications. The article proposes to use an approach based on the processing of a sequence of video segments in temporal and intra-frame spaces using the detection and elimination of new structural-statistical and structural and contextual regularities in the spectral basis. Here, to build new approaches to coding video segments, including informative video segments, it is proposed to use the theoretical foundations of non-equilibrium positional coding in spectral space.

As video-streaming services have expanded and improved, cloud-based video has evolved into a necessary feature of any successful business for reaching internal and external audiences. In this article, video streaming over distributed storage is considered where the video segments are encoded using an erasure code for better reliability. We consider a representative system architecture for a realistic (typical) content delivery network (CDN). Given multiple parallel streams/link between each server and the edge router, we need to determine, for each client request, the subset of servers to stream the video, as well as one of the parallel streams from each chosen server. To have this scheduling, this article proposes a two-stage probabilistic scheduling. The selection of video quality is also chosen with a certain probability distribution that is optimized in our algorithm. With these parameters, the playback time of video segments is determined by characterizing the download time of each coded chunk for each video segment. Using the playback times, a bound on the moment generating function of the stall duration is used to bound the mean stall duration. Based on this, we formulate an optimization problem to jointly optimize the convex combination of mean stall duration and average video quality for all requests, where the two-stage probabilistic scheduling, video quality selection, bandwidth split among parallel streams, and auxiliary bound parameters can be chosen. This non-convex problem is solved using an efficient iterative algorithm. Based on the offline version of our proposed algorithm, an online policy is developed where servers selection, quality, bandwidth split, and parallel streams are selected in an online manner. Experimental results show significant improvement in QoE metrics for cloud-based video as compared to the considered baselines.

Identifying anomalies in multidimensional data sets is an important yet challenging task in many real-world applications. A special case arises when anomalies are occluded in a small subset of attributes. We propose a new subspace analysis approach, called agglomerative attribute grouping (AAG), that searches for subspaces composed of highly correlative (in the general sense) attributes. Such correlations among attributes can better reflect the behavior of normal observations and hence, can be used to improve the identification of abnormal data samples. The proposed AAG algorithm relies on a generalized multiattribute measure (derived from information theory measures over attributes’ partitions) for evaluating the “information distance” among various subsets of attributes. To determine the set of subspaces, AAG applies a variation of the well-known agglomerative clustering algorithm with the proposed measure as the underlying distance function, whereas in contrast to existing methods, AAG does not require any tuning of parameters. Finally, the set of informative subspaces can be used to improve subspace-based analytical tasks, such as anomaly detection, novelty detection, forecasting, and clustering. Extensive evaluation over real-world data sets demonstrates that (i) in the vast majority of cases, AAG outperforms both classical and state-of-the-art subspace analysis methods when used in anomaly and novelty detection ensembles; (ii) it often generates fewer subspaces with fewer attributes each, thus resulting in faster training times for the anomaly and novelty detection ensemble; and (iii) the generated subspaces can also be useful in other analytical tasks, such as clustering and forecasting. History: Kwok-Leung Tsui served as the senior editor for this article. Funding: This research was partially supported by the Israeli Ministry of Economy (METRO 450 Consortium within the frame of MAGNET program) as well as by the Koret foundation grant for Smart Cities and Digital Living 2030. Data Ethics &amp; Reproducibility Note: The code capsule is available on Code Ocean at https://codeocean.com/capsule/2526218/tree/v1 and in the e-Companion to this article (available at https://doi.org/10.1287/ijds.2023.0027 ).

For the performance evaluation of future wireless communication systems, such as the fourth generation wireless networks, traffic traces of realistic services are needed. Multiple description coding (MDC) is gaining a lot of interest lately and is a viable solution to provide robust video services over single or multi hop wireless networks and MDC introduces more flexibility for network coding. Furthermore it has the capability to support heterogeneous terminals as they are accepted to be used in 4G wireless networks. By means of twelve well known video sequences, in different video formats, we generate the frame size traces and evaluate the multiple description coding characteristics. In addition to that we highlight the expected overhead due to the underlying RTP/UDP/IP protocol suite. As an objective quality measurement at the application layer, we investigate the video quality in dependency of lost and error-prone descriptors. This allows researcher to convert the network losses of their network models directly into video quality values. This step makes the work unique as single layer coded video streams would always need further postprocessing to retrieve the video quality.