Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks130 from public data
- Transformers in medical imaging: A survey1,338
This survey surveys the use of Transformers in medical image segmentation, detection, classification, restoration, synthesis, registration, clinical report generation, and other tasks and develops taxonomy for each application.
- The Visual Object Tracking VOT2016 Challenge Results776
The Visual Object Tracking challenge VOT2016 goes beyond its predecessors by introducing a new semi-automatic ground truth bounding box annotation methodology and extending the evaluation system with the no-reset experiment.
- Generative Cooperative Learning for Unsupervised Video Anomaly Detection236
This work proposes a novel unsupervised Generative Cooperative Learning (GCL) approach for video anomaly detection that exploits the low frequency of anomalies towards building a cross-supervision between a generator and a discriminator.
- Visual Object Tracking with Discriminative Filters and Siamese Networks: A Survey and Outlook224
This survey presents a systematic and thorough review of more than 90 DCFs and Siamese trackers, based on results in nine tracking benchmarks, and distinguishes and comprehensively review the shared as well as specific open research challenges in both these tracking paradigms.
- Mask-Guided Attention Network for Occluded Pedestrian Detection223
A novel mask-guided attention network is introduced that emphasizes on visible pedestrian regions while suppressing the occluded ones by modulating full body features and it is empirically demonstrated that coarse-level segmentation annotations provide reasonable approximation to their dense pixel-wise counterparts.
- Deep Contextual Attention for Human-Object Interaction Detection138
This work proposes a contextual attention framework for human-object interaction detection that leverages context by learning contextually-aware appearance features for human and object instances and adaptively selects relevant instance-centric context information to highlight image regions likely to contain human- object interactions.
- 72
- Pushing the boundaries of audiovisual word recognition using Residual Networks and LSTMs66
A deep learning architecture for lipreading and audiovisual word recognition is presented, which combines Residual Networks equipped with spatiotemporal input layers and Bidirectional LSTMs and further analysis on the utility of target word boundaries is provided.
- Alzheimer's disease: 3‐Dimensional MRI texture for prediction of conversion from mild cognitive impairment64
The main objective was to determine whether MRI texture could be used to predict conversion of MCI to AD.
- AnimalWeb: A Large-Scale Hierarchical Dataset of Annotated Animal Faces61
A large-scale, hierarchical annotated dataset of animal faces, featuring 22.4K faces from 350 diverse species and 21 animal orders across biological taxonomy is introduced, showing its challenging nature, unique attributes and present definite prospects for novel, adaptive, and generalized face-oriented CV algorithms.
- Improving Single Domain-Generalized Object Detection: A Focus on Diversification and Alignment53
This work demonstrates that by carefully selecting a set of augmentations, a base detector can outperform existing methods for single domain generalization by a good margin and introduces a method to align detections from multiple views, considering both classification and localization outputs.
- Fusion and Orthogonal Projection for Improved Face-Voice Association53
This work hypothesize that enriched feature representation coupled with an effective yet efficient supervision is necessary in realizing a discriminative joint embedding space for improved face-voice association and proposes a light-weight, plug-and-play mechanism that exploits the complementary cues in both modalities to form enriched fused embeddings and clusters them based on their identity labels via orthogonality constraints.
- Realistic and Efficient Face Swapping: A Unified Approach with Diffusion Models52
This work proposes to reframe the face-swapping task as a self-supervised, train-time inpainting problem, enhancing the identity transfer while blending with the target image, and introduces a mask shuffling technique during inpainting training, which allows for a so-called universal model for swapping.
- 51
- ThinkGeo: Evaluating Tool-Augmented Agents for Remote Sensing Tasks49
ThinkGeo, an agentic benchmark designed to evaluate LLM-driven agents on remote sensing tasks via structured tool use and multi-step planning, provides the first extensive testbed for evaluating how tool-enabled LLMs handle spatial reasoning in remote sensing.
- EcoTaskSched: a hybrid machine learning approach for energy-efficient task scheduling in IoT-based fog-cloud environments49
A novel ML-based EcoTaskSched model is proposed, which leverages deep learning for energy-efficient task scheduling in fog-cloud networks and is successfully applied to fog-cloud computing environments, increasing task handling efficiency and reducing energy consumption while maintaining the required QoS parameters.
- Synergy between Face Alignment and Tracking via Discriminative Global Consensus Optimization43
This work describes the very first, to the best of the knowledge, synergistic approach between detection and tracking which completely eliminates drifting from face tracking, and does not merely perform tracking-by-detection.
- The Canadian ALS Neuroimaging Consortium (CALSNIC) - a multicentre platform for standardized imaging and clinical studies in ALS38
The Canadian ALS Neuroimaging Consortium (CALSNIC) was founded to develop and evaluate advanced MRI-based biomarkers that delineate biological heterogeneity, track disease progression, and predict survival in a large and heterogeneous sample of ALS patients.
- Single-branch Network for Multimodal Training35
This work proposes a novel single-branch network capable of learning discriminative representation of unimodal as well as multimodal tasks without changing the network and demonstrates the superiority of this network over the existing methods in a wide range of experiments.
- Multiclass Confidence and Localization Calibration for Object Detection35
This paper proposes a new train-time technique for calibrating modern object detection methods that is capable of jointly calibrating multiclass confidence and box localization by leveraging their predictive uncertainties.
- 34
- 32
- 32
- TRIC-track: Tracking by Regression with Incrementally Learned Cascades32
This paper proposes to use cascaded regression with incremental learning to track generic objects without any prior knowledge of an object's structure or appearance by replacing local matching of an appearance model by direct prediction of the displacement between local image patches and part locations.
- Rich Semantics Improve Few-Shot Learning30
This work creates a bottleneck visual feature which is then used to generate language descriptions of the classes as an auxiliary task during training, and develops a Transformer based forward and backward encoding mechanism to relate visual and semantic tokens that can encode intricate relationships between the two modalities.
- Safety and feasibility of same-day discharge for catheter ablation of atrial fibrillation: a systematic review and meta-analysis27
In carefully selected AF patients undergoing CA, same-day discharge strategy is feasible and safe and there are no significant differences in post-discharge 30-day hospital visits, post- Discharge vascular complications, and other safety outcomes.
- Reliability of 3D texture analysis: A multicenter MRI study of the brain27
Texture analysis is an image‐analysis technique that detects complex intervoxel statistical patterns that has shown potential in detecting cerebral degeneration not perceptible to the human eye in many neurological disorders.
- 26
- Domain Adaptive Object Detection via Balancing Between Self-Training and Adversarial Learning25
This paper develops a technique to quantify predictive uncertainty on class assignments and bounding-box predictions and shows that this approach outperforms existing state-of-the-art methods with noticeable margins.
- Outcomes of robotic versus laparoscopic-assisted surgery in patients with rectal cancer: a systematic review and meta-analysis24
RARS is safe and feasible for rectal cancer patients and may be superior or equivalent to Laparoscopic-assisted rectal surgery, but high-standard, large-scale trials are required to determine the best approach.
- Bayesian prior models for vehicle make and model recognition23
A simple but powerful probabilistic framework for vehicle type recognition that requires just a single representative car image in the database to recognize any incoming test image exhibiting strong appearance variations, as expected in outdoor image capture e.g. illumination, scale etc.
- Why Not Use Your Textbook? Knowledge-Enhanced Procedure Planning of Instructional Videos22
Experimental evaluations reveal that KEPP attains superior, state-of-the-art results while requiring only mini-mal supervision, and harnesses a probabilistic procedural knowledge graph ex-tracted from training data, effectively acting as a compre-hensive textbook for the training domain.
- Distinct patterns of progressive gray and white matter degeneration in amyotrophic lateral sclerosis21
The findings suggest that the characteristic longitudinal gray matter pathology in ALS is the progressive involvement of frontotemporal regions rather than a worsening pathology within the motor cortex, and that phenotypic variability is associated with distinct progressive spatial pathology.
- Generalizing to Unseen Domains in Diabetic Retinopathy Classification20
A simple and effective domain generalization (DG) approach that achieves self-distillation in vision transformers (ViT) via a novel prediction softening mechanism that is capable of delivering improved calibration performance than other methods, showing its suitability for safety-critical applications, including health-care.
- A Probabilistic Framework for Patch based Vehicle Type Recognition.18
A novel patch selection criterion that automatically learns a discriminative patch set corresponding to vehicle regions for each vehicle class during training phase is proposed and it has shown highly competitive results on segmented car dataset and promising results on introduced dataset of cars.
- 17
- Mode-Guided Feature Augmentation for Domain Generalization17
This paper proposes a simple andcient DG approach to augment source domain(s) by hypothesizing the existence of favourable correlation between the source and target domain’s major modes of variation, and upon exploring those modes in the source domain the authors can realize meaningful alterations to background, appearance, pose and texture of object classes.
- 17
- 16
- AV-Deepfake1M++: A Large-Scale Audio-Visual Deepfake Benchmark with Real-World Perturbations16
This paper proposes AV-Deepfake1M++, an extension of the AV-Deepfake1M having 2 million video clips with diversified manipulation strategy and audio-visual perturbation, and believes that this dataset will play a pivotal role in facilitating research in Deepfake domain.
- HM: Hybrid Masking for Few-Shot Segmentation16
This work compensates for the loss of fine-grained spatial details in FM technique by investigating and leveraging a complementary basic input masking method.
- A Synopsis of FAME 2024 Challenge: Associating Faces with Voices in Multilingual Environments15
The Face-Voice Association in Multilingual Environments (FAME) 2024 Challenge, held at ACM Multimedia 2024, focuses on establishing face-voice association to analyze the impact of multiple languages on the verification process.
- SynFER: Towards Boosting Facial Expression Recognition With Synthetic Data13
This work introduces SynFER (Synthesis of Facial Expressions with Refined Control), a novel synthetic framework for synthesizing facial expression image data based on high-level textual descriptions as well as more fine-grained and precise control through facial action units.
- Multimodal Cardiac Imaging Revisited by Artificial Intelligence: An Innovative Way of Assessment or Just an Aid?13
This review explores the integration of artificial intelligence in multimodal cardiac imaging, tracing its evolution from early X-rays to contemporary techniques such as CT, MRI, and nuclear imaging, promising enhanced diagnostics and personalized patient care.
- Multi-Modal Medical Image Fusion for Non-Small Cell Lung Cancer Classification12
An innovative integration of multi-modal data is introduced, synthesizing fused medical imaging with clinical health records and genomic data, which has the potential to transform NSCLC diagnostics, facilitating earlier detection and more effective treatment planning and, ultimately, leading to superior patient outcomes in lung cancer care.
- Uninterrupted versus interrupted direct oral anticoagulation for catheter ablation of atrial fibrillation: A systematic review and meta‐analysis11
To evaluate the safety of uninterrupted versus interrupted direct oral anticoagulation (DOAC) for patients undergoing catheter ablation of atrial fibrillation (AF).
- DPA: Dual Prototypes Alignment for Unsupervised Adaptation of Vision-Language Models10
DPA introduces the concept of dual prototypes, acting as distinct classifiers, along with the convex combination of their outputs, thereby leading to accurate pseudo-label construction, and addresses visual-textual misalignment by aligning textual prototypes with image prototypes to further improve the adaptation performance.
- 10
- 10
- 10
- Effects of Perinatal Stroke on Executive Functioning and Mathematics Performance in Children10
Examination of executive functioning, math performance, and visuospatial processing skills of children with perinatal stroke found difficulties with executive function (inhibition) may contribute to math difficulties in this population.
- Comparison of Unicompartmental Knee Arthroplasty Versus High Tibial Osteotomy for Medial Knee Osteoarthritis: An Updated Meta‐Analysis of 56,000 Patients9
UKA is superior to HTO, offering lower revision rates, reduced postoperative pain, fewer complications, and better functional scores.
- OSLoPrompt: Bridging Low-Supervision Challenges and Open-Set Domain Generalization in CLIP9
This work proposes OSLoPrompt, an advanced prompt-learning framework for CLIP with two core innovations, and introduces a domainagnostic prompt-learning mechanism that integrates adaptable domain-specific cues and visually guided semantic attributes through a novel cross-attention module to improve outlier rejection during inference.
- 9
- A robust background subtraction algorithm for motion based video scene segmentation in embedded platforms9
Robust background subtraction algorithm to segment motion based video scene in embedded platforms to extract illumination invariant scene features and then combine them efficiently into the background reference frame is presented.
- 8
- Modality Invariant Multimodal Learning to Handle Missing Modalities: A Single-Branch Approach8
This work proposes a modality invariant multimodal learning method, which is less susceptible to the impact of missing modalities, and consists of a single-branch network sharing weights across multiple modalities to learn inter-modality representations to maximize performance as well as robustness to missing modalities.
- 8
- 8
- 7
- Comparative Efficacy and Safety of Different Tenecteplase Doses With Alteplase in Acute Ischemic Stroke: A Systematic Review With Pairwise and Network Meta‐Analysis to Determine the Optimal Dose6
This meta‐analysis evaluates the safety and efficacy of various TNK doses compared to alteplase, integrating recent randomized controlled trials (RCTs) and employing a frequentist network meta‐analysis to identify the optimal dose while addressing existing evidence gaps.
- Pose-Guided Self-Training with Two-Stage Clustering for Unsupervised Landmark Discovery6
A ULD algorithm based on diffusion features using self-training and clustering which also outperforms prior methods by notable margins is developed and a new proxy task based on generating latent pose codes is introduced, resulting in a significant performance improvement.
- Discrete Cycle-Consistency Based Unsupervised Deep Graph Matching6
The method does not require ground truth correspondences between keypoint pairs, and is self-supervised by enforcing consistency of matchings between images of the same object category, which makes it compatible with arbitrary network architectures and combinatorial solvers.
- Learning Branched Fusion and Orthogonal Projection for Face-Voice Association6
A light-weight, plug-and-play mechanism that exploits the complementary cues in both modalities to form enriched fused embeddings and clusters them based on their identity labels via orthogonality constraints is proposed and instantiate in a two-stream network.
- 6
- 6
- Towards Robust Multimodal Domain Generalization via Modality-Domain Joint Adversarial Training5
Modality-Domain Joint Adversarial Training is proposed, a unified framework that addresses challenges through two key innovations: a tri-discriminator adversarial module that mitigates domain biases in both modality-specific and multimodal representations, while suppressing modality-heterogeneous patterns in the representation space.
- 5
- FrogDogNet: Fourier Frequency Retained Visual Prompt Output Guidance for Domain Generalization of CLIP in Remote Sensing5
FrogDogNet is proposed, a novel prompt learning framework integrating Fourier frequency filtering and self-attention to improve RS scene classification and domain generalization, and highlights the effectiveness of frequency-based invariant feature retention in generalization.
- 5
- 5
- 5
- 5
- Unsupervised Deep Graph Matching Based on Cycle Consistency4
The method does not require ground truth correspondences between keypoint pairs, and is self-supervised by enforcing consistency of matchings between images of the same object category, which makes it compatible with arbitrary network architectures and combinatorial solvers.
- 4
- MTS: A Multiple Temporal Scale Tracker Handling Occlusion and Abrupt Motion Variation4
Improved and rich prior distribution, formed by the models learned and applied over multiple temporal scales, further makes the proposed method robust to complex target motion through covering relatively large search space with reduced sampling effort.
- 4
- 4
- 3
- Leveraging Cycle-Consistent Anchor Points for Self-Supervised RGB-D Registration3
This work uses cycle-consistent keypoints as salient points to enforce spatial coherence constraints during matching, improving correspondence accuracy and introducing a novel pose block that combines a GRU recurrent unit with transformation synchronization, blending historical and multi-view data.
- Executive behavior and functional abilities in children with perinatal stroke and the associated caregiver impact3
Caregivers of children with perinatal stroke who experience limitations in performing daily activities should be more closely monitored for adverse impact and be provided the necessary support and education to alleviate the associated guilt.
- Towards Fine-Grained Adaptation of CLIP via a Self-Trained Alignment Score2
This work introduces Fine-grained Alignment and Interaction Refinement (FAIR), an innovative approach that dynamically aligns localized image features with descriptive language embeddings through a set of Class Description Anchors, which enables the definition of a Learned Alignment Score (LAS), which incorporates CDA as an adaptive classifier, facilitating cross-modal interactions to improve self-training in unsupervised adaptation.
- FCC: Fully Connected Correlation for One-Shot Segmentation2
Fully Connected Correlation (Fully Connected Correlation) is introduced which integrates pixel-level correlations between support and query features, capturing associations that reveal target-specific patterns and correspondences in both same-layers and cross-layers.
- 2
- Efficacy and safety of laquinimod versus placebo in relapsing-remitting multiple sclerosis: a systematic review and meta-analysis of randomized controlled trials2
Findings support the efficacy of laquinimod in managing RRMS but necessitate careful monitoring during treatment.
- Clinical Features of Multidrug-Resistant Gram-Negative Bacteremia: A Comparative Study of Cancer and Non-Cancer Patients2
Given the rising prevalence of MDR-GNB pathogens, future research is needed for prompt appropriate antibacterial therapy based on risk stratification and enhanced antimicrobial stewardship programs as these are critical for high-risk cancer patients.
- Towards PerSense++: Advancing Training-Free Personalized Instance Segmentation in Dense Images2
PerSense, an end-to-end, training-free, and model-agnostic one-shot framework for Personalized instance Segmentation in dense images, and PerSense-D, a dedicated benchmark for personalized segmentation in dense images demonstrate that PerSense++ outperforms existing methods in dense settings.
- Knowledge and Attitude of the General Population Toward E-cigarette Use in Khyber Pakhtunkhwa: A Cross-Sectional Study2
Views were similar across genders, but most users were young adults from higher socioeconomic backgrounds, likely due to greater exposure to marketing and cultural norms, as the first study of its kind in Khyber Pakhtunkhwa highlights the need for public education and targeted regulation.
- Association of DNA damage response pathway genes with rheumatoid arthritis risks: a case-control study2
It is shown that DNA damage response pathway genes can act as efficient/specific diagnostic markers for said disease and lay a solid foundation for further research based on targeted metabolomics/genomics, which may lead to the development of more effective treatment strategies in the future.
- Prevalence of Post-traumatic Stress Disorder in Adult Burn Patients: A Study From Urban Tertiary Care Hospitals in Pakistan2
An alarmingly high prevalence of PTSD was found in burn patients, and was more prevalent in females than in males, and a greater percentage of retirement-age and middle-aged adults were diagnosed with PTSD.
- 2
- 2
- Cancer as an Individual Risk Factor for Heart Failure: A Review of Literature2
Emerging evidence suggesting a direct interaction between cancer and HF is highlighted, with studies indicating an elevated risk of mortality from cancer in patients with HF and vice versa, paving the way for multifaceted preventive and therapeutic approaches.
- 2
- 1
- Leveraging 2D Priors and SDF Guidance for Dynamic Urban Scene Rendering1
This work presents a novel approach that integrates Signed Distance Functions (SDFs) with 3D Gaussian Splatting to create a more robust object representation by harnessing the strengths of both methods, and enhances the geometric accuracy of 3D Gaussian splatting and improves deformation modeling within the SDF, resulting in a more adaptable and precise representation.
- TLAC: Two-Stage LMM Augmented CLIP for Zero-Shot Classification1
Two simple yet effective training-free approaches are introduced, Single-stage LMM Augmented CLIP (SLAC) and Two-stage LMM Augmented CLIP (TLAC), that leverages powerful Large Multimodal Models (LMMs) for image classification that leverages the capabilities of pretrained LMMs, allowing for seamless adaptation to diverse datasets and domains without the need for additional training.
- Artificial Intelligence-Driven Business Intelligence: Machine Learning Techniques for Financial Market Analysis1
A comprehensive AI-driven financial analysis model that features highly effective, data-driven decision-making tools for policymakers, financial analysts, and investors and emphasises the potential of AI to improve financial intelligence by enhancing predictive accuracy, risk management, and decision-making.
- Transfer Learning: A Key Approach to Fault Prediction and Extraction in Deep Learning1
A novel approach using 3D deep convolutional neural networks (DCNN) with transfer learning from six dip lines with transfer learning from six dip lines to incorporate local structural knowledge is proposed, demonstrating its robustness and effectiveness in complex fault systems.
- 1
- 1
- Attention Based Simple Primitives for Open-World Compositional Zero-Shot Learning1
This study is exploring Open-World Compositional Zero Shot Learning (OW-CZSL) in this study, where the test space encompasses all potential combinations of attributes and objects and the proposed model, Attention-based Simple Primitives (ASP), demonstrates competitive performance, achieving results comparable to the state-of-the-art.
- 1
- Application of Neural Networks to Scheduling Problem including Transportation Time1
This paper presents an approach for finding an optimal schedule of n-jobs and m-machines flowshop scheduling problem involving transportation time between jobs by using neural networks.
- Efficacy and safety of daratumumab-based quadruplet therapy vs non-daratumumab-based triplet therapy in newly diagnosed multiple myeloma patients: A GRADE assessed systematic review and meta-analysis of randomized controlled trials–
The meta-analysis shows higher efficacy yet higher adverse events in Daratumumab-based quadruplet therapy for NDMM patients (TE, and TIE).
- –
- –
- –
- Post‐Dengue Deep Vein Thrombosis Complicated by Pulmonary Embolism: A Case Report–
The case of a 25‐year‐old previously healthy male who presented with high‐grade fever, myalgia, and headache and was diagnosed with dengue fever based on positive NS1 antigen testing demonstrates the paradoxical coexistence of thrombosis and bleeding risk in dengue fever.
- –
- –
- –
- –
- Expanding the Phenotypic and Genotypic Spectrum of Postaxial Polydactyly: A Study of Four Consanguineous Pakistani Families–
This study has broadened the phenotypic and allelic spectrum of the genes associated with isolated PAP and strengthened the role of these genes in regulating limb development.
- Cervical cancer mortality trends and disparities in women of the United States: A two-decade retrospective analysis.–
The mortality due to cervical cancer has decreased over the past 2 decades, but there are significant disparities in mortality in certain groups especially elderly females, Black or African American females, and females in the southern region.
- –
- –
- Palbociclib plus endocrine therapy or GnRH analogs versus capecitabine or placebo for HR+/HER2- metastatic breast cancer: A systematic review and meta-analysis.–
Palbociclib, when used in combination with ET, showed improvement in PFS and ORR in HR+/HER2- metastatic breast cancer patients as compared to capecitabine or placebo but demonstrated no superiority in terms of overall MS and median RD.
- –
- –
- –
- Unlocking the Hidden Pathway: A Rare Encounter With Right Coronary Artery Fistula to Coronary Sinus and Right Ventricle–
A 62-year-old male with a complex medical history, including hypertension, atrial fibrillation, and heart failure, who presented with exertional chest pain and palpitations was presented with a significant coronary artery fistula, deemed clinically significant due to its potential to cause myocardial ischemia.
- Efficacy and Safety of Daratumumab-Based Quadruplet Therapy Vs Non-Daratumumab-Based Triplet Therapy in Newly Diagnosed Multiple Myeloma Patients: A Systematic Review and Meta-Analysis–
Primary efficacy outcomes focused on PFS and stringent complete response (sCR) and secondary outcomes included overall response rate (ORR), complete response (CR), complete response or better (≥CR), very good partial response rate (VGPR), very good partial response rate or better (VGPR), negative MRD, and safety profiles.
- –
- 216. A Comparison of Multidrug-Resistant Gram-negative Bacteremia in Cancer and Non-Cancer Patients–
The data showed that the common source of MDR-GNB remains urinary tract infection in both groups and the bacterial epidemiology is comparable between the two groups.
- COMPUTER VISION AND AI IN MEDICAL IMAGING: ENHANCING DIAGNOSTIC ACCURACY IN RADIOLOGY–
This article explores the transformative potential of AI-driven image analysis systems, particularly convolutional neural networks (CNNs), in detecting and classifying abnormalities such as tumors, fractures, and pulmonary diseases.
- Quantifying Cerebral Degeneration in Amyotrophic Lateral Sclerosis Using Texture Analysis of Multimodal MRI (4560)–
Texture analysis on T1-weighted and FLAIR MRI can potentially identify changes consistent with degeneration in the precentral gyri and corticospinal tracts in patients with ALS.
- Abstract 13720: Safety of Same-day Discharge for Atrial Fibrillation Ablation: A Systematic Review and Meta-analysis–
Same day discharge after atrial fibrillation ablation is safe and has the potential to be a cost-effective strategy, and Randomized trials are needed to validate these results.
- Visual tracking over multiple temporal scales–
A visual tracker operating over multiple temporal scales is proposed that is capable of handling occlusions and abrupt motion variations and a search method is generalized for multiple competing hypotheses in visual tracking, and a new motion model selection criterion is proposed.
- –
Publication data from OpenAlex, with missing venues and authors filled in from Crossref; citation counts are the higher of OpenAlex and Semantic Scholar, last synced 2026-10-15. One-sentence summaries under some papers are written by Semantic Scholar’s model. Citation counts may be lower than on Google Scholar, which indexes more sources.
Report an error
Wrong papers, two people merged into one, or a profile that should not be here? Tell us and we will fix or hide it.