arXiv subjects

Lin Wang

Publications and source records attributed to Lin Wang.

At least 199 records · Page 11Linked to original sources

arXiv2024

DAP-LED: Learning Degradation-Aware Priors with CLIP for Joint Low-light Enhancement and Deblurring

Autonomous vehicles and robots often struggle with reliable visual perception at night due to the low illumination and motion blur caused by the long exposure time of RGB cameras. Existing methods address this challenge by sequentially connecting the off-the-shelf pretrained low-light enhancement and deblurring models. Unfortunately, these methods often lead to noticeable artifacts (\eg, color distortions) in the over-exposed regions or make it hardly possible to learn the motion cues of the dark regions. In this paper, we interestingly find vision-language models, \eg, Contrastive Language-Image Pretraining (CLIP), can comprehensively perceive diverse degradation levels at night. In light of this, we propose a novel transformer-based joint learning framework, named DAP-LED, which can jointly achieve low-light enhancement and deblurring, benefiting downstream tasks, such as depth estimation, segmentation, and detection in the dark. The key insight is to leverage CLIP to adaptively learn the degradation levels from images at night. This subtly enables learning rich semantic information and visual representation for optimization of the joint tasks. To achieve this, we first introduce a CLIP-guided cross-fusion module to obtain multi-scale patch-wise degradation heatmaps from the image embeddings. Then, the heatmaps are fused via the designed CLIP-enhanced transformer blocks to retain useful degradation information for effective model optimization. Experimental results show that, compared to existing methods, our DAP-LED achieves state-of-the-art performance in the dark. Meanwhile, the enhanced results are demonstrated to be effective for three downstream tasks. For demo and more results, please check the project page: \url{https://vlislab22.github.io/dap-led/}.

cs.CV↗

arXiv2024

Robo-GS: A Physics Consistent Spatial-Temporal Model for Robotic Arm with Hybrid Representation

Real2Sim2Real plays a critical role in robotic arm control and reinforcement learning, yet bridging this gap remains a significant challenge due to the complex physical properties of robots and the objects they manipulate. Existing methods lack a comprehensive solution to accurately reconstruct real-world objects with spatial representations and their associated physics attributes. We propose a Real2Sim pipeline with a hybrid representation model that integrates mesh geometry, 3D Gaussian kernels, and physics attributes to enhance the digital asset representation of robotic arms. This hybrid representation is implemented through a Gaussian-Mesh-Pixel binding technique, which establishes an isomorphic mapping between mesh vertices and Gaussian models. This enables a fully differentiable rendering pipeline that can be optimized through numerical solvers, achieves high-fidelity rendering via Gaussian Splatting, and facilitates physically plausible simulation of the robotic arm's interaction with its environment using mesh-based methods. The code,full presentation and datasets will be made publicly available at our website https://robostudioapp.com

cs.RO↗

arXiv2024

G-NeLF: Memory- and Data-Efficient Hybrid Neural Light Field for Novel View Synthesis

Following the burgeoning interest in implicit neural representation, Neural Light Field (NeLF) has been introduced to predict the color of a ray directly. Unlike Neural Radiance Field (NeRF), NeLF does not create a point-wise representation by predicting color and volume density for each point in space. However, the current NeLF methods face a challenge as they need to train a NeRF model first and then synthesize over 10K views to train NeLF for improved performance. Additionally, the rendering quality of NeLF methods is lower compared to NeRF methods. In this paper, we propose G-NeLF, a versatile grid-based NeLF approach that utilizes spatial-aware features to unleash the potential of the neural network's inference capability, and consequently overcome the difficulties of NeLF training. Specifically, we employ a spatial-aware feature sequence derived from a meticulously crafted grid as the ray's representation. Drawing from our empirical studies on the adaptability of multi-resolution hash tables, we introduce a novel grid-based ray representation for NeLF that can represent the entire space with a very limited number of parameters. To better utilize the sequence feature, we design a lightweight ray color decoder that simulates the ray propagation process, enabling a more efficient inference of the ray's color. G-NeLF can be trained without necessitating significant storage overhead and with the model size of only 0.95 MB to surpass previous state-of-the-art NeLF. Moreover, compared with grid-based NeRF methods, e.g., Instant-NGP, we only utilize one-tenth of its parameters to achieve higher performance. Our code will be released upon acceptance.

cs.CV↗

arXiv2024

Quantum Effects on Cosmic Scales as an Alternative to Dark Matter and Dark Energy

The spin-torsion theory is a gauge theory approach to gravity that expands upon Einstein's general relativity (GR) by incorporating the spin of microparticles. In this study, we further develop the spin-torsion theory to examine spherically symmetric and static gravitational systems that involve free-falling macroscopic particles. We posit that the quantum spin of macroscopic matter becomes noteworthy at cosmic scales. We further assume that the Dirac spinor and Dirac equation adequately capture all essential physical characteristics of the particles and their associated processes. A crucial aspect of our approach involves substituting the constant mass in the Dirac equation with a scale function, allowing us to establish a connection between quantum effects and the scale of gravitational systems. This mechanism ensures that the quantum effect of macroscopic matter is scale-dependent and diminishes locally, a phenomenon not observed in microparticles. For any given matter density distribution, our theory predicts an additional quantum term, the quantum potential energy (QPE), within the mass expression. The QPE induces time dilation and distance contraction, and thus mimics a gravitational well. When applied to cosmology, the QPE serves as a counterpart to the cosmological constant introduced by Einstein to balance gravity in his static cosmological model. The QPE also offers a plausible explanation for the origin of Hubble redshift (traditionally attributed to the universe's expansion). The predicted luminosity distance--redshift relation aligns remarkably well with SNe Ia data from the cosmological sample of SNe Ia. In the context of galaxies, the QPE functions as the equivalent of dark matter. The predicted circular velocities align well with rotation curve data from the SPARC (Spitzer Photometry and Accurate Rotation Curves database) sample.

gr-qc↗

arXiv2024

EvLight++: Low-Light Video Enhancement with an Event Camera: A Large-Scale Real-World Dataset, Novel Method, and More

Event cameras offer significant advantages for low-light video enhancement, primarily due to their high dynamic range. Current research, however, is severely limited by the absence of large-scale, real-world, and spatio-temporally aligned event-video datasets. To address this, we introduce a large-scale dataset with over 30,000 pairs of frames and events captured under varying illumination. This dataset was curated using a robotic arm that traces a consistent non-linear trajectory, achieving spatial alignment precision under 0.03mm and temporal alignment with errors under 0.01s for 90% of the dataset. Based on the dataset, we propose \textbf{EvLight++}, a novel event-guided low-light video enhancement approach designed for robust performance in real-world scenarios. Firstly, we design a multi-scale holistic fusion branch to integrate structural and textural information from both images and events. To counteract variations in regional illumination and noise, we introduce Signal-to-Noise Ratio (SNR)-guided regional feature selection, enhancing features from high SNR regions and augmenting those from low SNR regions by extracting structural information from events. To incorporate temporal information and ensure temporal coherence, we further introduce a recurrent module and temporal loss in the whole pipeline. Extensive experiments on our and the synthetic SDSD dataset demonstrate that EvLight++ significantly outperforms both single image- and video-based methods by 1.37 dB and 3.71 dB, respectively. To further explore its potential in downstream tasks like semantic segmentation and monocular depth estimation, we extend our datasets by adding pseudo segmentation and depth labels via meticulous annotation efforts with foundation models. Experiments under diverse low-light scenes show that the enhanced results achieve a 15.97% improvement in mIoU for semantic segmentation.

cs.CV↗

arXiv2024

Centering the Value of Every Modality: Towards Efficient and Resilient Modality-agnostic Semantic Segmentation

Fusing an arbitrary number of modalities is vital for achieving robust multi-modal fusion of semantic segmentation yet remains less explored to date. Recent endeavors regard RGB modality as the center and the others as the auxiliary, yielding an asymmetric architecture with two branches. However, the RGB modality may struggle in certain circumstances, e.g., nighttime, while others, e.g., event data, own their merits; thus, it is imperative for the fusion model to discern robust and fragile modalities, and incorporate the most robust and fragile ones to learn a resilient multi-modal framework. To this end, we propose a novel method, named MAGIC, that can be flexibly paired with various backbones, ranging from compact to high-performance models. Our method comprises two key plug-and-play modules. Firstly, we introduce a multi-modal aggregation module to efficiently process features from multi-modal batches and extract complementary scene information. On top, a unified arbitrary-modal selection module is proposed to utilize the aggregated features as the benchmark to rank the multi-modal features based on the similarity scores. This way, our method can eliminate the dependence on RGB modality and better overcome sensor failures while ensuring the segmentation performance. Under the commonly considered multi-modal setting, our method achieves state-of-the-art performance while reducing the model parameters by 60%. Moreover, our method is superior in the novel modality-agnostic setting, where it outperforms prior arts by a large margin of +19.41% mIoU

cs.CV↗

arXiv2024

Unsignalized Intersection Management Strategy for Mixed Autonomy Traffic Streams

With the rapid development of connected and automated vehicles (CAVs) and intelligent transportation infrastructure, CAVs, connected human-driven vehicles (CHVs), and un-connected human-driven vehicles (HVs) will coexist on the roads in the future for a long time. This paper comprehensively considers the different traffic characteristics of CHVs, CAVs, and HVs, and systemically investigates the unsignalized intersection management strategy from the upper decision-making level to the lower execution level. The unsignalized intersection management strategy consists of two parts: the heuristic priority queues based right of way allocation (HPQ) algorithm and the vehicle planning and control algorithm. In the HPQ algorithm, a vehicle priority management model considering the difference between CAVs, CHVs, and HVs, is built to design the right of way management for different types of vehicles. In the lower level for vehicle planning and control algorithm, different control modes of CAVs are designed according to the upper-level decision made by the HPQ algorithm. Moreover, the vehicle control execution is realized by the model predictive controller combined with the geographical environment constraints and the unsignalized intersection management strategy. The proposed strategy is evaluated by simulations, which show that the proposed intersection management strategy can effectively reduce travel time and improve traffic efficiency. Results show that the proposed method can decrease the average travel time by 5% to 65% for different traffic flows compared with the comparative methods. The intersection management strategy captures the real-world balance between efficiency and safety for future intelligent traffic systems.

eess.SY↗

arXiv2024

Elite360M: Efficient 360 Multi-task Learning via Bi-projection Fusion and Cross-task Collaboration

360 cameras capture the entire surrounding environment with a large FoV, exhibiting comprehensive visual information to directly infer the 3D structures, e.g., depth and surface normal, and semantic information simultaneously. Existing works predominantly specialize in a single task, leaving multi-task learning of 3D geometry and semantics largely unexplored. Achieving such an objective is, however, challenging due to: 1) inherent spherical distortion of planar equirectangular projection (ERP) and insufficient global perception induced by 360 image's ultra-wide FoV; 2) non-trivial progress in effectively merging geometry and semantics among different tasks to achieve mutual benefits. In this paper, we propose a novel end-to-end multi-task learning framework, named Elite360M, capable of inferring 3D structures via depth and surface normal estimation, and semantics via semantic segmentation simultaneously. Our key idea is to build a representation with strong global perception and less distortion while exploring the inter- and cross-task relationships between geometry and semantics. We incorporate the distortion-free and spatially continuous icosahedron projection (ICOSAP) points and combine them with ERP to enhance global perception. With a negligible cost, a Bi-projection Bi-attention Fusion module is thus designed to capture the semantic- and distance-aware dependencies between each pixel of the region-aware ERP feature and the ICOSAP point feature set. Moreover, we propose a novel Cross-task Collaboration module to explicitly extract task-specific geometric and semantic information from the learned representation to achieve preliminary predictions. It then integrates the spatial contextual information among tasks to realize cross-task fusion. Extensive experiments demonstrate the effectiveness and efficacy of Elite360M.

cs.CV↗

arXiv2024

GoodSAM++: Bridging Domain and Capacity Gaps via Segment Anything Model for Panoramic Semantic Segmentation

This paper presents GoodSAM++, a novel framework utilizing the powerful zero-shot instance segmentation capability of SAM (i.e., teacher) to learn a compact panoramic semantic segmentation model, i.e., student, without requiring any labeled data. GoodSAM++ addresses two critical challenges: 1) SAM's inability to provide semantic labels and inherent distortion problems of panoramic images; 2) the significant capacity disparity between SAM and the student. The `out-of-the-box' insight of GoodSAM++ is to introduce a teacher assistant (TA) to provide semantic information for SAM, integrated with SAM to obtain reliable pseudo semantic maps to bridge both domain and capacity gaps. To make this possible, we first propose a Distortion-Aware Rectification (DARv2) module to address the domain gap. It effectively mitigates the object deformation and distortion problem in panoramic images to obtain pseudo semantic maps. We then introduce a Multi-level Knowledge Adaptation (MKA) module to efficiently transfer the semantic information from the TA and pseudo semantic maps to our compact student model, addressing the significant capacity gap. We conduct extensive experiments on both outdoor and indoor benchmark datasets, showing that our GoodSAM++ achieves a remarkable performance improvement over the state-of-the-art (SOTA) domain adaptation methods. Moreover, diverse open-world scenarios demonstrate the generalization capacity of our GoodSAM++. Last but not least, our most lightweight student model achieves comparable performance to the SOTA models with only 3.7 million parameters.

cs.CV↗

arXiv2024

A General Framework to Boost 3D GS Initialization for Text-to-3D Generation by Lexical Richness

Text-to-3D content creation has recently received much attention, especially with the prevalence of 3D Gaussians Splatting. In general, GS-based methods comprise two key stages: initialization and rendering optimization. To achieve initialization, existing works directly apply random sphere initialization or 3D diffusion models, e.g., Point-E, to derive the initial shapes. However, such strategies suffer from two critical yet challenging problems: 1) the final shapes are still similar to the initial ones even after training; 2) shapes can be produced only from simple texts, e.g., "a dog", not for lexically richer texts, e.g., "a dog is sitting on the top of the airplane". To address these problems, this paper proposes a novel general framework to boost the 3D GS Initialization for text-to-3D generation upon the lexical richness. Our key idea is to aggregate 3D Gaussians into spatially uniform voxels to represent complex shapes while enabling the spatial interaction among the 3D Gaussians and semantic interaction between Gaussians and texts. Specifically, we first construct a voxelized representation, where each voxel holds a 3D Gaussian with its position, scale, and rotation fixed while setting opacity as the sole factor to determine a position's occupancy. We then design an initialization network mainly consisting of two novel components: 1) Global Information Perception (GIP) block and 2) Gaussians-Text Fusion (GTF) block. Such a design enables each 3D Gaussian to assimilate the spatial information from other areas and semantic information from texts. Extensive experiments show the superiority of our framework of high-quality 3D GS initialization against the existing methods, e.g., Shap-E, by taking lexically simple, medium, and hard texts. Also, our framework can be seamlessly plugged into SoTA training frameworks, e.g., LucidDreamer, for semantically consistent text-to-3D generation.

cs.CV↗

arXiv2024

Dimeric Drug Polymeric Micelles with Acid-Active Tumor Targeting and FRET-indicated Drug Release

Trans-activating transcriptional activator (TAT), a cell-penetrating peptide, has been extensively used for facilitating cellular uptake and nuclear targeting of drug delivery systems. However, the positively charged TAT peptide usually strongly interacts with serum components and undergoes substantial phagocytosis by the reticuloendothelial system, causing a short blood circulation in vivo. In this work, an acid-active tumor targeting nanoplatform DA-TAT-PECL was developed to effectively inhibit the nonspecific interactions of TAT in the bloodstream. 2,3-dimethylmaleic anhydride (DA) was first used to convert the TAT amines to carboxylic acid, the resulting DA-TAT was further conjugated to get DA-TAT-PECL. After self-assembly into polymeric micelles, they were capable of circulating in the physiological condition for a long time and promoting cell penetration upon accumulation at the tumor site and de-shielding the DA group. Moreover, camptothecin (CPT) was used as the anticancer drug and modified into a dimer (CPT)2-ss-Mal, in which two CPT molecules were connected by a reduction-labile maleimide thioether bond. The FRET signal between CPT and maleimide thioether bond was monitored to visualize the drug release process and effective targeted delivery of antitumor drugs was demonstrated. This pH/reduction dual-responsive micelle system provides a new platform for high fidelity cancer therapy.

q-bio.TO↗

arXiv2024

Revisit Event Generation Model: Self-Supervised Learning of Event-to-Video Reconstruction with Implicit Neural Representations

Reconstructing intensity frames from event data while maintaining high temporal resolution and dynamic range is crucial for bridging the gap between event-based and frame-based computer vision. Previous approaches have depended on supervised learning on synthetic data, which lacks interpretability and risk over-fitting to the setting of the event simulator. Recently, self-supervised learning (SSL) based methods, which primarily utilize per-frame optical flow to estimate intensity via photometric constancy, has been actively investigated. However, they are vulnerable to errors in the case of inaccurate optical flow. This paper proposes a novel SSL event-to-video reconstruction approach, dubbed EvINR, which eliminates the need for labeled data or optical flow estimation. Our core idea is to reconstruct intensity frames by directly addressing the event generation model, essentially a partial differential equation (PDE) that describes how events are generated based on the time-varying brightness signals. Specifically, we utilize an implicit neural representation (INR), which takes in spatiotemporal coordinate $(x, y, t)$ and predicts intensity values, to represent the solution of the event generation equation. The INR, parameterized as a fully-connected Multi-layer Perceptron (MLP), can be optimized with its temporal derivatives supervised by events. To make EvINR feasible for online requisites, we propose several acceleration techniques that substantially expedite the training process. Comprehensive experiments demonstrate that our EvINR surpasses previous SSL methods by 38% w.r.t. Mean Squared Error (MSE) and is comparable or superior to SoTA supervised methods. Project page: https://vlislab22.github.io/EvINR/.

cs.CV↗

arXiv2024

EventBind: Learning a Unified Representation to Bind Them All for Event-based Open-world Understanding

In this paper, we propose EventBind, a novel and effective framework that unleashes the potential of vision-language models (VLMs) for event-based recognition to compensate for the lack of large-scale event-based datasets. In particular, due to the distinct modality gap with the image-text data and the lack of large-scale datasets, learning a common representation space for images, texts, and events is non-trivial.Intuitively, we need to address two key challenges: 1) how to generalize CLIP's visual encoder to event data while fully leveraging events' unique properties, e.g., sparsity and high temporal resolution; 2) how to effectively align the multi-modal embeddings, i.e., image, text, and events. Accordingly, we first introduce a novel event encoder that subtly models the temporal information from events and meanwhile, generates event prompts for modality bridging. We then design a text encoder that generates content prompts and utilizes hybrid text prompts to enhance EventBind's generalization ability across diverse datasets.With the proposed event encoder, text encoder, and image encoder, a novel Hierarchical Triple Contrastive Alignment (HTCA) module is introduced to jointly optimize the correlation and enable efficient knowledge transfer among the three modalities. We evaluate various settings, including fine-tuning and few-shot on three benchmarks, and our EventBind achieves new state-of-the-art accuracy compared with the previous methods, such as on N-Caltech101 (+5.34% and +1.70%) and N-Imagenet (+5.65% and +1.99%) with fine-tuning and 20-shot settings, respectively. Moreover, our EventBind can be flexibly extended to the event retrieval task using text or image queries, showing plausible performance. Project page:https://vlislab22.github.io/EventBind/.

cs.CV↗

arXiv2024

SRFNet: Monocular Depth Estimation with Fine-grained Structure via Spatial Reliability-oriented Fusion of Frames and Events

Monocular depth estimation is a crucial task to measure distance relative to a camera, which is important for applications, such as robot navigation and self-driving. Traditional frame-based methods suffer from performance drops due to the limited dynamic range and motion blur. Therefore, recent works leverage novel event cameras to complement or guide the frame modality via frame-event feature fusion. However, event streams exhibit spatial sparsity, leaving some areas unperceived, especially in regions with marginal light changes. Therefore, direct fusion methods, e.g., RAMNet, often ignore the contribution of the most confident regions of each modality. This leads to structural ambiguity in the modality fusion process, thus degrading the depth estimation performance. In this paper, we propose a novel Spatial Reliability-oriented Fusion Network (SRFNet), that can estimate depth with fine-grained structure at both daytime and nighttime. Our method consists of two key technical components. Firstly, we propose an attention-based interactive fusion (AIF) module that applies spatial priors of events and frames as the initial masks and learns the consensus regions to guide the inter-modal feature fusion. The fused feature are then fed back to enhance the frame and event feature learning. Meanwhile, it utilizes an output head to generate a fused mask, which is iteratively updated for learning consensual spatial priors. Secondly, we propose the Reliability-oriented Depth Refinement (RDR) module to estimate dense depth with the fine-grained structure based on the fused features and masks. We evaluate the effectiveness of our method on the synthetic and real-world datasets, which shows that, even without pretraining, our method outperforms the prior methods, e.g., RAMNet, especially in night scenes. Our project homepage: https://vlislab22.github.io/SRFNet.

cs.CV↗

arXiv2024

Data On the Go: Seamless Data Routing for Intermittently-Powered Battery-Free Sensing

The rising demand for sustainable IoT has promoted the adoption of battery-free devices intermittently powered by ambient energy for sensing. However, the intermittency poses significant challenges in sensing data collection. Despite recent efforts to enable one-to-one communication, routing data across multiple intermittently-powered battery-free devices, a crucial requirement for a sensing system, remains a formidable challenge. This paper fills this gap by introducing Swift, which enables seamless data routing in intermittently-powered battery-free sensing systems. Swift overcomes the challenges posed by device intermittency and heterogeneous energy conditions through three major innovative designs. First, Swift incorporates a reliable node synchronization protocol backed by number theory, ensuring successful synchronization regardless of energy conditions. Second, Swift adopts a low-latency message forwarding protocol, allowing continuous message forwarding without repeated synchronization. Finally, Swift features a simple yet effective mechanism for routing path construction, enabling nodes to obtain the optimal path to the sink node with minimum hops. We implement Swift and perform large-scale experiments representing diverse realworld scenarios. The results demonstrate that Swift achieves an order of magnitude reduction in end-to-end message delivery time compared with the state-of-the-art approaches for intermittentlypowered battery-free sensing systems.

cs.NI↗

arXiv2024

A Tale of Two Scales: Reconciling Horizontal and Vertical Scaling for Inference Serving Systems

Inference serving is of great importance in deploying machine learning models in real-world applications, ensuring efficient processing and quick responses to inference requests. However, managing resources in these systems poses significant challenges, particularly in maintaining performance under varying and unpredictable workloads. Two primary scaling strategies, horizontal and vertical scaling, offer different advantages and limitations. Horizontal scaling adds more instances to handle increased loads but can suffer from cold start issues and increased management complexity. Vertical scaling boosts the capacity of existing instances, allowing for quicker responses but is limited by hardware and model parallelization capabilities. This paper introduces Themis, a system designed to leverage the benefits of both horizontal and vertical scaling in inference serving systems. Themis employs a two-stage autoscaling strategy: initially using in-place vertical scaling to handle workload surges and then switching to horizontal scaling to optimize resource efficiency once the workload stabilizes. The system profiles the processing latency of deep learning models, calculates queuing delays, and employs different dynamic programming algorithms to solve the joint horizontal and vertical scaling problem optimally based on the workload situation. Extensive evaluations with real-world workload traces demonstrate over $10\times$ SLO violation reduction compared to the state-of-the-art horizontal or vertical autoscaling approaches while maintaining resource efficiency when the workload is stable.

cs.DC↗

arXiv2024

DreamScene: 3D Gaussian-based Text-to-3D Scene Generation via Formation Pattern Sampling

Text-to-3D scene generation holds immense potential for the gaming, film, and architecture sectors. Despite significant progress, existing methods struggle with maintaining high quality, consistency, and editing flexibility. In this paper, we propose DreamScene, a 3D Gaussian-based novel text-to-3D scene generation framework, to tackle the aforementioned three challenges mainly via two strategies. First, DreamScene employs Formation Pattern Sampling (FPS), a multi-timestep sampling strategy guided by the formation patterns of 3D objects, to form fast, semantically rich, and high-quality representations. FPS uses 3D Gaussian filtering for optimization stability, and leverages reconstruction techniques to generate plausible textures. Second, DreamScene employs a progressive three-stage camera sampling strategy, specifically designed for both indoor and outdoor settings, to effectively ensure object-environment integration and scene-wide 3D consistency. Last, DreamScene enhances scene editing flexibility by integrating objects and environments, enabling targeted adjustments. Extensive experiments validate DreamScene's superiority over current state-of-the-art techniques, heralding its wide-ranging potential for diverse applications. Code and demos will be released at https://dreamscene-project.github.io .

cs.CV↗

arXiv2024

OphNet: A Large-Scale Video Benchmark for Ophthalmic Surgical Workflow Understanding

Surgical scene perception via videos is critical for advancing robotic surgery, telesurgery, and AI-assisted surgery, particularly in ophthalmology. However, the scarcity of diverse and richly annotated video datasets has hindered the development of intelligent systems for surgical workflow analysis. Existing datasets face challenges such as small scale, lack of diversity in surgery and phase categories, and absence of time-localized annotations. These limitations impede action understanding and model generalization validation in complex and diverse real-world surgical scenarios. To address this gap, we introduce OphNet, a large-scale, expert-annotated video benchmark for ophthalmic surgical workflow understanding. OphNet features: 1) A diverse collection of 2,278 surgical videos spanning 66 types of cataract, glaucoma, and corneal surgeries, with detailed annotations for 102 unique surgical phases and 150 fine-grained operations. 2) Sequential and hierarchical annotations for each surgery, phase, and operation, enabling comprehensive understanding and improved interpretability. 3) Time-localized annotations, facilitating temporal localization and prediction tasks within surgical workflows. With approximately 285 hours of surgical videos, OphNet is about 20 times larger than the largest existing surgical workflow analysis benchmark. Code and dataset are available at: https://minghu0830.github.io/OphNet-benchmark/.

cs.CV↗