Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks52 from public data
- 101
- NTIRE 2022 Challenge on Efficient Super-Resolution: Methods and Results94
The NTIRE 2022 challenge was to super-resolve an input image with a magnification factor of ×4 based on pairs of low and corresponding high resolution images and the aim was to design a network for single image super-resolution that achieved improvement of efficiency measured according to several metrics.
- Multi-stage segmentation of white matter hyperintensity, cortical and lacunar infarcts63
An automated multi-stage segmentation approach for labeling the WMH, CI, and LI using multi-modal MR images and the evaluations against the manual labeling and ARWMC visual grading suggest that the algorithm provides reasonable segmentation accuracy.
- Towards Compact Single Image Super-Resolution via Contrastive Self-distillation61
A novel contrastive self-distillation (CSD) framework to simultaneously compress and accelerate various off-the-shelf SR models to improve the quality of SR images and PSNR/SSIM via explicit knowledge transfer is proposed.
- Griffin: Towards a Graph-Centric Relational Database Foundation Model43
Evaluated on large-scale, heterogeneous, and temporal graphs extracted from RDBs across various domains, Griffin demonstrates superior or comparable performance to individually trained models, excels in low-data scenarios, and shows strong transferability with similarity and diversity in pretraining across new datasets and tasks, highlighting its potential as a universally applicable foundation model for RDBs.
- 42
- Slot-VAE: Object-Centric Scene Generation with Slot Attention31
The Slot-VAE is proposed, a generative model that integrates slot attention with the hierarchical VAE framework for object-centric structured scene generation and outperforms slot representation-based generative baselines in terms of sample quality and scene structure accuracy.
- 23
- 18
- 16
- 14
- 14
- Do We Really Need Curated Malicious Data for Safety Alignment in Multi-modal Large Language Models?11
Experiments show that, without the need for labor-intensive collection of high-quality malicious data, model safety can still be significantly improved, as long as a specific fraction of rejection data exists in the finetuning set, indicating the security alignment is not lost but rather obscured during multi-modal pretraining or instruction finetuning.
- Towards Eliminating Hard Label Constraints in Gradient Inversion Attacks8
This work is the first to initiate a novel algorithm to simultaneously recover the ground-truth augmented label and the input feature of the last fully-connected layer from single-input gradients, and provide a necessary condition for any analytical-based label recovery methods.
- SALT: A Flexible Semi-Automatic Labeling Tool for General LiDAR Point Clouds with Cross-Scene Adaptability and 4D Consistency7
This work proposes a novel zero-shot learning paradigm, termed data alignment, which transforms LiDAR data into pseudo-images by aligning with the training distribution of vision foundation models, and designs a 4D-consistent prompting strategy and 4D non-maximum suppression module to enhance SAM2, ensuring high-quality, temporally consistent presegmentation.
- 7
- Sample Correlation for Fingerprinting Deep Face Recognition5
This work presents SAC-JC, a novel yet simple model stealing detection method based on SAmple Correlation that selects JPEG compressed samples as model inputs and calculates the correlation matrix among their model outputs, and extends it to object recognition datasets including Tiny-ImageNet and CIFAR10.
- 5
- 5
- 5
- Test-Time Immunization: A Universal Defense Framework Against Jailbreaks for (Multimodal) Large Language Models4
This work proposes a universal defense framework, termed Test-time IMmunization (TIM), which can adaptively defend against various jailbreak attacks in a self-evolving way and implements safety fine-tuning using the detected jailbreak instructions paired with refusal answers.
- Compositional Scene Understanding through Inverse Generative Modeling4
This work formulate scene understanding as an inverse generative modeling problem, where it is proposed to build this visual generative model compositionally from smaller models over pieces of a scene to infer scene structure from images substantially different than those seen during training.
- 4
- Container Application Migration Algorithm in Internet of Vehicles4
Simulation results show that the dynamic programming method proposed in this paper reduces the long-term average migration total cost by 33.88% and 24.53%, respectively, compared to the nearest selection method and the local optimal method.
- A META-ANALYTIC REVIEW OF SOCIAL MEDIA STUDIES4
The paper systematically reviewed social media studies published in Association for Information Systems (AIS) listed top 20 journals from 2009 to 2013 and formulate projections and recommendations for future social media Studies.
- Discovering anomaly on the basis of flow estimation of alert feature distribution3
The idea of filtering irrelevant alerts based on alert volumes is extended and the flow estimation of abrupt changes in feature distribution caused by anomalies is formulated, by computing Kullback-Leibler distance of alert feature values under observation in comparison with a reference distribution.
- 2
- 2
- Garbage Sorting System Based on Composite Layer CNN and Multi-Robots2
A garbage sorting system based on convolution neural network and robots, which combines the features of three different network layers and predicts the location and angle information of the target objects at the same time, which is called composite layer is proposed.
- Research on SLAM Drift Reduction Mechanism Based on Point Cloud Segmentation Semantic Information2
This paper combines the semantic segmentation of scenes with Simultaneous localization and Mapping (SLAM) technology to build a three-dimensional semantic map, which can reduce the system’s translation drift error by 18.8%.
- 2
- 2
- GILT: An LLM-Free, Tuning-Free Graph Foundational Model for In-Context Learning1
GILT introduces a novel token-based framework for in-context learning (ICL) on graphs, reframing classification tasks spanning node, edge and graph levels in a unified framework designed to operate on generic numerical features.
- Research on the Online Monitoring and Fault Early Warning System of Infrared Thermography for Wind Power Equipment Based on Deep Learning Algorithm1
The research results show that the research results can provide an intelligent solution for the condition monitoring of wind power equipment, effectively reduce the risk of unplanned downtime, and improve the operation and maintenance efficiency and economic benefits.
- 1
- 1
- 1
- A Compression and Recognition Joint Model for Structured Video Surveillance Storage1
The proposed joint model stores the video in a structured format, allowing high-level vision tasks such as structured storage, content analysis, and intelligent retrieval, and saves nearly 28% of the number of computing resources.
- 1
- RETRACTED: A method of multi-criteria set recognition based on deep feature representation1
A method to improve the performance of the re identification using depth learning and multiple metric ensembles and a large number of experiments show that this method can effectively improve the recognition performance compared to the state-of-the-art methods.
- SecTube: SGX-Based Trusted Transmission System1
A novel approach called SecTube is presented to protect the data safety in transmission that utilizes Intel’s new security technology SGX to give user application a safer execution environment and the design and implementation of enclave socket are presented.
- Efficient large-scale graph data optimization for intelligent video surveillance1
A comprehensive survey of recent studies to solve the problem of topology learning, Object appearance modeling and global activity understanding sparse camera network is presented and some of the current open Research issues are discussed.
- A Secret Sharing Scheme to Recover Secret Unnecessary Calculating1
A secret sharing threshold scheme based on direct partitio ning and combining of set which avoids any computation is presented which extends to general access structures.
- –
- –
- –
- –
- –
- –
- –
- –
- –
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 2026-10-14. 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.