Is this you? Claim this profile to correct it, add a bio and choose the work people see first.
Claim this profileWorks152 from public data
- All in One: Exploring Unified Video-Language Pre-Training258
This work introduces an end-to-end VLP model, namely all-in-one Transformer, that embeds raw video and textual signals into joint representations using a unified backbone architecture and introduces a novel and effective token rolling operation to encode temporal representations from video clips in a non-parametric manner.
- 222
- Inhibition of highly productive HIV-1 infection in T cells, primary human macrophages, microglia, and astrocytes by Sargassum fusiforme172
This is the first report demonstrating S. fusiforme to be a potent inhibitor of highly productive HIV-1 infection and replication in T cells, in primary human macrophages, microglia, and astrocytes.
- Self-Monitoring and Self-Delivery of Photosensitizer-Doped Nanoparticles for Highly Effective Combination Cancer Therapyin Vitroandin Vivo169
Overall, the carrier-free DDS is able to achieve chemotherapy, photodynamic therapy, and real-time self-monitoring of the release and distribution of the nanomedicine (Cur and H2TPyP) and the as-prepared NPs show high cancer therapeutic efficiency both in vitro and in vivo.
- CLIP-Event: Connecting Text and Images with Event Structures159
A contrastive learning framework to enforce vision-language pretraining models to comprehend events and associated argument (participant) roles is proposed, which takes advantage of text information extraction technologies to obtain event structural knowledge, and utilizes multiple prompt functions to contrast difficult negative descriptions by manipulating event structures.
- 159
- DMC-Net: Generating Discriminative Motion Cues for Fast Compressed Video Action Recognition155
A lightweight generator network, which reduces noises in motion vectors and captures fine motion details, achieving a more Discriminative Motion Cue (DMC) representation, is proposed which obtains high accuracy close to that of using flow and runs two orders of magnitude faster than using optical flow at inference time.
- Deep Variational Metric Learning117
This paper proposes a deep variational metric learning (DVML) framework to explicitly model the intra-class variance and disentangle the intra -class invariance, namely, the class centers, and can simultaneously generate discriminative samples to improve robustness.
- Organic electrochemical transistor arrays for real-time mapping of evoked neurotransmitter release in vivo114
The OECT-array is an active sensor with intrinsic amplification capability, allowing real-time and direct readout of transient CA-NT release with a sensitivity of nanomolar range and a temporal resolution of several milliseconds, and reveals a complex cross-talk between the mesolimbic and the nigrostriatal pathways.
- Learning To Recognize Procedural Activities with Distant Supervision109
This paper proposes to automatically identify steps in instructional videos by leveraging the distant supervision of a textual knowledge base that includes detailed descriptions of the steps needed for the execution of a wide variety of complex activities, and demonstrates that video models trained to recognize these automatically-labeled steps yield a representation that achieves superior generalization performance on four downstream tasks.
- Tetherless near-infrared control of brain activity in behaving animals using fully implantable upconversion microdevices96
This study presents an all-optical method for tetherless remote control of neural activity using fully implantable micro-devices based on upconversion technology and shows that remotely applied NIR illumination is able to reliably trigger spiking activity in rat brains.
- Object-aware Video-language Pre-training for Retrieval92
Object-aware Transformers is presented, an object-centric approach that extends video-language transformer to incorporate object representations to leverage the bounding boxes and object tags to guide the training process.
- Multiplexed Optogenetic Stimulation of Neurons with Spectrum‐Selective Upconversion Nanoparticles84
The UCNP‐based optrode is implanted into the brain of live rodents to achieve all‐optical remote activation of brain tissues in mammalian animals and it is believed that this proof‐of‐concept study opens up completely new applications of upconversion materials for regulating physiological functions, especially in neuroscience research.
- Cryogenic 3D printing of heterogeneous scaffolds with gradient mechanical strengths and spatial delivery of osteogenic peptide/TGF-β1 for osteochondral tissue regeneration82
This proof-of-concept study provides a facile way to produce integrated osteochondral scaffolds for concurrently directing rBMSC osteogenic/chondrogenic differentiation at different regions.
- VX2TEXT: End-to-End Learning of Video-Based Text Generation From Multimodal Inputs79
Unlike prior encoder-only models, the proposed VX2TEXT network includes an autoregressive decoder to generate open-ended text from the multimodal embeddings fused by the language encoder, rendering the approach fully generative and makes it directly applicable to different "video+x to text" problems without the need to design specialized network heads for each task.
- Context-Gated Convolution77
This work proposes one novel Context-Gated Convolution (CGC) to explicitly modify the weights of convolutional layers adaptively under the guidance of global context, which can better extract representative local patterns and compose discriminative features in CNNs.
- Remote modulation of neural activities via near-infrared triggered release of biomolecules77
This platform is based on microscale composite hydrogels that incorporate polypyrrole (PPy) nanoparticles as photo-thermal transducers and is triggered by near-infrared-light (NIR) irradiation, and is shown to work similarly well in vivo, demonstrating remote control of brain activity without any genetic modifications.
- High-throughput mapping of brain-wide activity in awake and drug-responsive vertebrates75
A microfluidic array system ("Fish-Trap") that enables high-throughput mapping of brain-wide activity in awake larval zebrafish and tests a set of neurotoxin peptides and obtained new insights into how to exploit neurotoxin derivatives as therapeutic agents.
- A novel CNN-DDPG based AI-trader: Performance and roles in business operations70
A novel “Reinforcement Learning” (RL) framework based AI-trader is built that is shown to outperform other methods with the use of real stock-index future data and an actor-critic RL algorithm called DDPG is adopted to find the optimal policy.
- 68
- RESIN: A Dockerized Schema-Guided Cross-document Cross-lingual Cross-media Information Extraction and Event Tracking System65
A new information extraction system that can automatically construct temporal event graphs from a collection of news documents from multiple sources, multiple languages, and multiple data modalities is presented.
- Supervised Masked Knowledge Distillation for Few-Shot Transformers62
A novel Supervised Masked Knowledge Distillation model (SMKD) for few-shot Transformers which incorporates label information into self-distillation frameworks and introduces the challenging task of masked patch tokens reconstruction across intra-class images.
- High-throughput brain activity mapping and machine learning as a foundation for systems neuropharmacology60
An in vivo drug screening strategy that combines high-throughput technology to generate large-scale brain activity maps to be used with machine learning to predict the therapeutic potential of clinically relevant drug leads is described.
- Palmitic Acid Is a Novel CD4 Fusion Inhibitor That Blocks HIV Entry and Infection57
Palmitic acid is introduced as a novel class of antiviral compound that binds directly to the CD4 receptor, blocking HIV-1 entry and infection and should lead to the development of PA analogs with greater potency against HIV-1 entry.
- Scheduling Tugboats in a Seaport56
This paper studies a tugboat scheduling problem in a busy port, where incoming and outgoing vessels frequently require the assistance of tugboats, but the number of available tugboats is limited, and develops a novel iterative solution method which combines Lagrangian relaxation and Benders decomposition.
- Controlled nanoparticle release from stable magnetic microbubble oscillations56
The ability of MMBs to release their particle cargo on demand under stable oscillation is reported, suggesting opportunities for targeted drug delivery through/into tissues that are not easily diffused through or penetrated.
- An upconversion nanoprobe operating in the first biological window56
A core-shell-shell nanostructure is described to realize photon upconversion in the first biological spectral window where the absorption of water and the biological specimen is minimal and developed as optical bioprobes that offer sensitive biodetection in the presence of tissue wrapping.
- NeuroArray: A Universal Interface for Patterning and Interrogating Neural Circuitry with Single Cell Resolution55
The NeuroArray is a universal patterning platform designed for, but not limited to neuron cells that can be readily interfaced with other interrogation modalities for high-throughput drug testing and for building neuron culture based live computational devices.
- 51
- 46
- An implantable ultrasound-powered device for the treatment of brain cancer using electromagnetic fields45
In vitro and in vivo experiments confirmed the significant therapeutic effect of the UP-TTD, with ~58% inhibition on growth rate of clinical tumor cells and ~78% reduction of cancer area in tumor-bearing rats.
- RESIN-11: Schema-guided Event Prediction for 11 Newsworthy Scenarios39
This work introduces RESIN-11, a new schema-guided event extraction&prediction framework that can be applied to a large variety of newsworthy scenarios and builds a demo website based on the dockerized system and schema library.
- MuMuQA: Multimedia Multi-Hop News Question Answering via Cross-Media Knowledge Extraction and Grounding37
A new QA evaluation benchmark with 1,384 questions over news articles that require cross- media grounding of objects in images onto text, and introduces a novel multimedia data augmentation framework, based on cross-media knowledge extraction and synthetic question-answer generation, to automatically augment data that can provide weak supervision for this task.
- Wireless Intelligent Patch for Closed‐loop In Situ Wound Management35
A smart wound patch system designed for wireless closed‐loop and in‐situ wound management and enables personalized treatment is presented, expanding the arsenal of tools available for effective wound management and enables personalized treatment.
- Joint Multimedia Event Extraction from Video and Article35
This work introduces the new task of Video MultiMedia Event Extraction (Video M2E2) and proposes the first self-supervised multimodal event coreference model that can determine coreference between video events and text events without any manually annotated pairs and introduces the first multimodals transformer which extracts structured event information jointly from both videos and text documents.
- 35
- Co-selection mechanism for bacterial resistance to major chemical pollutants in the environment34
The necessity of studying the co-selection mechanism of bacteria against major chemical pollutants against major chemical pollutants is emphasised, emphasising the necessity of studying the co-selection mechanism of bacteria against major chemical pollutants, and lays a solid theoretical foundation for conducting risk assessment of bacterial resistance.
- Fish Capsules: A System for High‐Throughput Screening of Combinatorial Drugs34
A novel fish capsule system that combines automated zebrafish encapsulating technology and droplet microarray strategy for in vivo functional screening of mono/polytherapies is described, and the prompt trapping of fish capsules in the aqueous arrays is successfully demonstrate.
- Mean First Passage Time and Stochastic Resonance in a Transcriptional Regulatory System with Non-Gaussian Noise31
The mean first passage time (MFPT) in a phenomenological gene transcriptional regulatory model with non-Gaussian noise is analytically investigated based on the singular perturbation technique and it is found that bounded non- Gaussian noise inhibits the transition between different concentrations of protein, while heavy-tailed non-gaussian noise accelerates the transition.
- Video-Text Pre-training with Learned Regions for Retrieval30
This work proposes a simple yet effective module for video-text representation learning, namely RegionLearner, which can take into account the structure of objects during pre-training on large-scale video- text pairs and is much more computationally efficient.
- Towards Fast Adaptation of Pretrained Contrastive Models for Multi-channel Video-Language Retrieval28
It is found that discrete text tokens coupled with a pretrained contrastive text model yields the best performance, which can even outperform state-of-the-art on the iVQA and How2QA datasets without additional training on millions of video-text data.
- 26
- 26
- 23
- 22
- 19
- 18
- 18
- 17
- 17
- 17
- 17
- 15
- 14
- 14
- 13
- 12
- 11
- 11
- 11
- 11
- 10
- 9
- 9
- 9
- 9
- 8
- 8
- 8
- 8
- 7
- 7
- 7
- 7
- 7
- 6
- 6
- 6
- Revitalize Region Feature for Democratizing Video-Language Pre-training of Retrieval6
This work revitalizes region features of sparsely sampled video clips to significantly reduce both spatial and temporal visual redundancy towards democratizing VLP research at the same time achieving state-of-the-art results.
- 6
- 6
- 6
- 5
- 5
- 4
- 4
- 4
- 4
- 4
- 4
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- 3
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 2
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- 1
- Consumers' Trust Maintenance: From the Perspective of Knowledge-Based and Identification-Based Trust1
- 1
- 1
- 1
- –
- –
- High‐Power Graphene Thermoacoustic Underwater Sound Source Assisted by SEBS Micromesh Matching Layer–
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
- –
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 2026-10-13. 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.