Event cameras are bio-inspired sensors that capture visual information through asynchronous brightness changes, offering distinct advantages including high temporal resolution and wide dynamic range. While prior research has investigated event-based 3D reconstruction for extreme scenarios, existing methods face inherent limitations and fail to fully exploit the unique characteristics of event data. In this paper, we present EvDiff3D, a novel two-stage 3D reconstruction framework that integrates event-based geometric constraints with an event-aware diffusion prior for appearance refinement. Our key insight lies in bridging the gap between physically grounded event-based reconstruction and data-driven appearance repair through a unified cyclical pipeline. In the first stage, we reconstruct a coarse 3D scene under supervision from event loss and event-based monocular depth constraints to preserve structural fidelity. The second stage fine-tunes an event-aware diffusion model based on a pretrained video diffusion model as a repair prior to enhance the appearance in under-constrained regions. Based on the diffusion model, our pipeline operates within a reconstruction-generation cycle that progressively refines both geometry and appearance using only event data. Extensive experiments on synthetic and real-world datasets demonstrate that EvDiff3D significantly outperforms existing methods in perceptual quality and structural consistency.

In this paper, we present Ev-iCRF, a novel self-supervised pipeline for high dynamic range (HDR) image reconstruction from a single-exposure low dynamic range (LDR) image, guided by asynchronous event streams generated by a bio-inspired event camera. The highlight of Ev-iCRF lies in its formulation of the inverse camera response function (iCRF) based on Event-LDR Correspondence. By leveraging the HDR properties of event data, the method enables direct iCRF estimation, offering a new perspective for event-guided HDR imaging. The pipeline is trained in a self-supervised manner using formulation-driven iCRF estimation loss and refinement loss, without the need for synchronized HDR supervision. Ev-iCRF adopts a two-stage coarse-to-fine reconstruction pipeline, allowing effective fusion of features from both LDR image and event data. The event information is used to optimize the iCRF, enabling accurate HDR reconstruction from LDR inputs. We evaluate Ev-iCRF on real-world datasets, and results show that it outperforms state-of-the-art methods in HDR reconstruction accuracy. Moreover, the reconstructed images demonstrate improved texture fidelity and structural detail.

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 0.03 mm and temporal alignment with errors under 0.01 s for 90% of the dataset. Based on the dataset, we propose EvLight++, a novel event-guided low-light video enhancement approach designed for robust performance in real-world scenarios. First, 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 ours 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.

Current text-to-avatar methods often rely on implicit representations (e.g., NeRF, SDF, and DMTet), leading to 3D content that artists cannot easily edit and animate in graphics software. This paper introduces a novel framework for generating stylized head avatars from text guidance, which leverages locally learnable mesh deformation and 2D diffusion priors to achieve high-quality digital assets for attribute-preserving manipulation. Given a template mesh, our method represents mesh deformation with perface Jacobians and adaptively modulates local deformation using a learnable vector field. This vector field enables anisotropic scaling while preserving the rotation of vertices, which can better express identity and geometric details. We also employ landmark- and contour-based regularization terms to balance the expressiveness and plausibility of generated head avatars from multiple views without relying on any specific shape prior. Our framework can generate realistic shapes and textures that can be further edited via text, while supporting seamless editing using the preserved attributes from the template mesh, such as 3DMM parameters, blendshapes, and UV coordinates. Extensive experiments demonstrate that our framework can generate diverse and expressive head avatars with high-quality meshes that artists can easily manipulate in 3D graphics software, facilitating downstream applications such as efficient asset creation and animation with preserved attributes.

Recently, Depth Anything Models (DAMs) [47], [48] - a type of depth foundation models – have demonstrated impressive zero-shot capabilities across diverse perspective images. Despite its success, it remains an open question regarding DAMs’ performance on panorama images that enjoy a large field-of-view (180° × 360°) but suffer from spherical distortions. To address this gap, we conduct an empirical analysis to evaluate the performance of DAMs on panoramic images and identify their limitations. For this, we undertake comprehensive experiments to assess the performance of DAMs from three key factors: panoramic representations, 360° camera positions for capturing scenarios, and spherical spatial transformations. This way, we reveal some key findings, e.g., DAMs are sensitive to spatial transformations. We then propose a semi-supervised learning (SSL) framework to learn a panoramic DAM, dubbed PanDA. Under the umbrella of SSL, PanDA first learns a teacher model by fine-tuning DAM through joint training on synthetic indoor and outdoor panoramic datasets. Then, a student model is trained using large-scale unlabeled data, leveraging pseudo-labels generated by the teacher model. To enhance PanDA’s generalization capability, Möbius transformation-based spatial augmentation (MTSA) is proposed to impose consistency regularization between the predicted depth maps from the original and spatially transformed ones. This subtly improves the student model’s robustness to various spatial transformations, even under severe distortions. Extensive experiments demonstrate that PanDA exhibits remarkable zero-shot capability across diverse scenes, and outperforms the data-specific panoramic depth estimation methods on two popular real-world benchmarks. Project page: https://caozidong.github.io/PanDA_Depth/.

Event cameras are bio-inspired sensors that output asynchronous and sparse event streams, instead of fixed frames. Benefiting from their distinct advantages, such as high dynamic range and high temporal resolution, event cameras have been applied to address 3D reconstruction, important for robotic mapping. Recently, neural rendering techniques, such as 3D Gaussian splatting (3DGS), have been shown successful in 3D reconstruction. However, it still remains under-explored how to develop an effective event-based 3DGS pipeline. In particular, as 3DGS typically depends on high-quality initialization and dense multiview constraints, a potential problem appears for the 3DGS optimization with events given its inherent sparse property. To this end, we propose a novel event-based 3DGS framework, named Elite-EvGS. Our key idea is to distill the prior knowledge from the off-the-shelf event-to-video (E2V) models to effectively reconstruct 3D scenes from events in a coarse-to-fine optimization manner. Specifically, to address the complexity of 3DGS initialization from events, we introduce a novel warm-up initialization strategy that optimizes a coarse 3DGS from the frames generated by E2V models and then incorporates events to refine the details. Then, we propose a progressive event supervision strategy that employs the window-slicing operation to progressively reduce the number of events used for supervision. This subtly relives the temporal randomness of the event frames, benefiting the optimization of local textural and global structural details. Experiments on the benchmark datasets demonstrate that Elite-EvGS can reconstruct 3D scenes with better textural and structural details. Meanwhile, our method yields plausible performance on the captured real-world data, including diverse challenging conditions, such as fast motion and low light scenes. For demo and more results, please check our project page.

With the rapid development of mobile and edge-integrated sensing technologies, 3D point clouds have emerged as a fundamental data modality for understanding and interacting with the physical world. They provide rich spatial and geometric information that enables autonomous driving, mobile robotics, and intelligent IoT devices to perceive and reason about their surroundings in real time. In particular, the development of edge-deployed sensors, such as LiDAR, depth cameras, and structured-light sensors, has made it feasible to capture and process 3D point clouds directly at the network edge, empowering low-latency perception and decision-making for safety-critical systems.

Blind face restoration (BFR) is a fundamental and challenging problem in computer vision. To faithfully restore high-quality (HQ) photos from poor-quality ones, recent research endeavors predominantly rely on facial image priors from the powerful pretrained text-to-image (T2I) diffusion models. However, such priors often lead to the incorrect generation of non-facial features and insufficient facial details, thus rendering them less practical for real-world applications. In this paper, we propose a novel framework, namely AuthFace that achieves highly authentic face restoration results by exploring a face-oriented generative diffusion prior. To learn such a prior, we first collect a dataset of 1.5K high-quality images, with resolutions exceeding 8K, captured by professional photographers. Based on the dataset, we then introduce a novel face-oriented restoration-tuning pipeline that fine-tunes a pretrained T2I model. Identifying key criteria of quality-first and photography-guided annotation, we involve the retouching and reviewing process under the guidance of photographers for high-quality images that show rich facial features. The photography-guided annotation system fully explores the potential of these high-quality photographic images. In this way, the potent natural image priors from pretrained T2I diffusion models can be subtly harnessed, specifically enhancing their capability in facial detail restoration. Moreover, to minimize artifacts in critical facial areas, such as eyes and mouth, we propose a time-aware latent facial feature loss to learn the authentic face restoration process. Extensive experiments on the synthetic and real-world BFR datasets demonstrate the superiority of our approach.

This paper presents PP-AICloud, a practical system prototype designed to enable privacy-preserving AI inference for cloud-based mobile services. Motivated by recent industry efforts, such as Apple's integration of Homomorphic Encryption (HE) for on-device intelligence, our work addresses the key limitations of existing privacy-preserving machine learning (PPML) solutions, i.e., high latency, bandwidth inefficiencies, and excessive on-device computation. PP-AICloud leverages edge nodes as an intermediate computing layer between mobile devices and centralized AICloud infrastructures, distributing the HE workflow across edge and cloud resources. By integrating HE with deep convolutional neural networks (CNNs), the system enables efficient and secure inference on encrypted user data without requiring decryption. Experimental results demonstrate that PP-AICloud achieves more than 91% accuracy on a real-world landmark recognition task with real-time latency of under 0.7 seconds. We demonstrate the capabilities of PP-AICloud through demo videos available at: Demo link.

Event cameras are bio-inspired sensors that capture intensity changes asynchronously with distinct advantages, such as high temporal resolution. Existing methods for event-based object/action recognition predominantly sample and convert event representation at every fixed temporal interval (or frequency). However, they are constrained to processing a limited number of event lengths and show poor frequency generalization, thus not fully leveraging the event's high temporal resolution. In this paper, we present our PASS framework, exhibiting superior capacity for spatiotemporal event modeling towards a larger number of event lengths and generalization across varying inference temporal frequencies. Our key insight is to learn adaptively encoded event features via the state space models (SSMs), whose linear complexity and generalization on input frequency make them ideal for processing high temporal resolution events. Specifically, we propose a Path-selective Event Aggregation and Scan (PEAS) module to encode events into features with fixed dimensions by adaptively scanning and selecting aggregated event presentation. On top of it, we introduce a novel Multi-faceted Selection Guiding (MSG) loss to minimize the randomness and redundancy of the encoded features during the PEAS selection process. Our method outperforms prior methods on five public datasets and shows strong generalization across varying inference frequencies with less accuracy drop (ours -8.62% v.s. -20.69% for the baseline). Moreover, our model exhibits strong long spatiotemporal modeling for a broader distribution of event length (1-10^9), precise temporal perception, and effective generalization for real-world scenarios. Code and checkpoints will be released upon acceptance.

In this paper, we address the challenging source-free unsupervised domain adaptation (SFUDA) for pinhole-to-panoramic semantic segmentation, given only a pinhole image pre-trained model (i.e., source) and unlabeled panoramic images (i.e., target). Tackling this problem is non-trivial due to three critical challenges: 1) semantic mismatches from the distinct Field-of-View (FoV) between domains, 2) style discrepancies inherent in the UDA problem, and 3) inevitable distortion of the panoramic images. To tackle these problems, we propose 360SFUDA++ that effectively extracts knowledge from the source pinhole model with only unlabeled panoramic images and transfers the reliable knowledge to the target panoramic domain. Specifically, we first utilize Tangent Projection (TP) as it has less distortion and meanwhile slits the equirectangular projection (ERP) to patches with fixed FoV projection (FFP) to mimic the pinhole images. Both projections are shown effective in extracting knowledge from the source model. However, as the distinct projections make it less possible to directly transfer knowledge between domains, we then propose Reliable Panoramic Prototype Adaptation Module (RPAM) to transfer knowledge at both prediction and prototype levels. RPAM selects the confident knowledge and integrates panoramic prototypes for reliable knowledge adaptation. Moreover, we introduce Cross-projection Dual Attention Module (CDAM), which better aligns the spatial and channel characteristics across projections at the feature level between domains. Both knowledge extraction and transfer processes are synchronously updated to reach the best performance. Extensive experiments on the synthetic and real-world benchmarks, including outdoor and indoor scenarios, demonstrate that our 360SFUDA++ achieves significantly better performance than prior SFUDA methods.

Source-free cross-modal knowledge transfer is a crucial yet challenging task, which aims to transfer knowledge from one source modality (e.g., RGB) to the target modality (e.g., depth or infrared) with no access to the task-relevant (TR) source data due to memory and privacy concerns. A recent attempt leverages the paired task-irrelevant (TI) data and directly matches the features from them to eliminate the modality gap. However, it ignores a pivotal clue that the paired TI data could be utilized to effectively estimate the source data distribution and better facilitate knowledge transfer to the target modality. To this end, we propose a novel yet concise framework to unlock the potential of paired TI data for enhancing source-free cross-modal knowledge transfer. Our work is buttressed by two key technical components. Firstly, to better estimate the source data distribution, we introduce a Task-irrelevant data-Guided Modality Bridging (TGMB) module. It translates the target modality data into the source-like images based on paired TI data and the guidance of the available source model to alleviate two key gaps: 1) inter-modality gap between the paired TI data; 2) intra-modality gap between TI and TR target data. We then propose a Task-irrelevant data-Guided Knowledge Transfer (TGKT) module that transfers knowledge from the source model to the target model by leveraging the paired TI data. Notably, due to the unavailability of labels for the TR target data and its less reliable prediction from the source model, our TGKT model incorporates a self-supervised pseudo-labeling approach to enable the target model to learn from its predictions. Extensive experiments show that our method achieves state-of-the-art performance on three datasets (RGB-to-depth and RGB-to-infrared).

Unsupervised visible-infrared person reidentification (UVI-ReID) has recently gained great attention due to its potential for enhancing human detection in diverse environments without labeling. Previous methods utilize intramodality clustering and cross-modality feature matching to achieve UVI-ReID. However, there exist two challenges: 1) noisy pseudo-labels might be generated in the clustering process and 2) the cross-modality feature alignment via matching the marginal distribution of visible and infrared modalities may misalign the different identities from the two modalities. In this article, we first conduct a theoretical analysis where an interpretable generalization upper bound is introduced. Based on the analysis, we then propose a novel unsupervised cross-modality person reidentification framework (PRAISE). Specifically, to address the first challenge, we propose a pseudo-label correction (PLC) strategy that utilizes a beta mixture model (BMM) to predict the probability of misclustering-based network's memory effect and rectifies the correspondence by adding a perceptual term to contrastive learning. Next, we introduce a modality-level alignment (MLA) strategy that generates paired visible-infrared latent features and reduces the modality gap by aligning the labeling function of visible and infrared features to learn identity-discriminative and modality-invariant features. Experimental results on two benchmark datasets demonstrate that our method achieves a state-of-the-art (SOTA) performance than the unsupervised visible-ReID methods.

Most consumer cameras use rolling shutter (RS) exposure, the captured videos often suffer from distortions (e.g., skew and jelly effect). Also, these videos are impeded by the limited bandwidth and frame rate, which inevitably affect the video streaming experience. In this paper, we excavate the potential of event cameras as they enjoy high temporal resolution. Accordingly, we propose a framework to recover the global shutter (GS) high frame rate (i.e., slow motion) video without RS distortion from an RS camera and event camera. One challenge is the lack of real-world datasets for supervised training. Therefore, we explore self-supervised learning with the key idea of estimating the displacement field-a non-linear and dense 3D spatiotemporal representation of all pixels during the exposure time. This allows for a mutual reconstruction between RS and GS frames and facilitates slow-motion video recovery. We then combine the input RS frames with the DF to map them to the GS frames (RS-to-GS). Given the under-constrained nature of this mapping, we integrate it with the inverse mapping (GS-to-RS) and RS frame warping (RS-to-RS) for self-supervision. We evaluate our framework via objective analysis (i.e., quantitative and qualitative comparisons on four datasets) and subjective studies (i.e., user study). The results show that our framework can recover slow-motion videos without distortion, with much lower bandwidth (94% drop) and higher inference speed ($ 16\; {\rm ms}/{\rm frame}$16 ms / frame ) under $32 \times$32× frame interpolation.

360 depth estimation has recently received great attention for 3D reconstruction owing to its omnidirectional field of view (FoV). Recent approaches are predominantly focused on cross-projection fusion with geometry-based reprojection: they fuse 360 images with equirectangular projection (ERP) and another projection type, e.g., cubemap projection to estimate depth with the ERP format. However, these methods suffer from 1) limited local receptive fields, making it hardly possible to capture large FoV scenes, and 2) prohibitive computational cost, caused by the complex cross-projection fusion module design. In this paper, we propose Elite360D, a novel framework that inputs the ERP image and icosahedron projection (ICOSAP) point set, which is undistorted and spatially continuous. Elite360D is superior in its capacity in learning a representation from a local-with-global perspective. With a flexible ERP image encoder, it includes an ICOSAP point encoder, and a Biprojection Bi-attention Fusion (B2F) module (totally ~1M parameters). Specifically, the ERP image encoder can take various perspective image-trained backbones (e.g., ResNet, Transformer) to extract local features. The point encoder extracts the global features from the ICOSAP. Then, the B2F module captures the semantic- and distance-aware dependencies between each pixel of the ERP feature and the entire ICOSAP feature set. Without specific backbone design and obvious computational cost increase, Elite360D outperforms the prior arts on several benchmark datasets.

Implicit neural representation (INR), in combination with geometric rendering, has recently been employed in real-time dense RGB-D SLAM. Despite active research endeavors being made, there lacks a unified protocol for fair evaluation, impeding the evolution of this area. In this work, we establish, to our knowledge, the first open-source benchmark framework to evaluate the performance of a wide spectrum of commonly used INRs and rendering functions for mapping and localization. The goal of our benchmark is to 1) gain an intuition of how different INRs and rendering functions impact mapping and localization and 2) establish a unified evaluation protocol w.r. t. the design choices that may impact the mapping and localization. With the framework, we conduct a large suite of experiments, offering various insights in choosing the INRs and geometric rendering functions: for example, the dense feature grid outperforms other INRs (e.g. tri-plane and hash grid), even when geometric and color features are jointly encoded for memory efficiency. To extend the findings into the practical scenario, a hybrid encoding strategy is proposed to bring the best of the accuracy and completion from the grid-based and decomposition-based INRs. We further propose explicit hybrid encoding for high-fidelity dense grid mapping to comply with the RGB-D SLAM system that puts the premise on robustness and computation efficiency.

Event camera has recently received much attention for low-light image enhancement (LIE) thanks to their distinct advantages, such as high dynamic range. However, current research is prohibitively restricted by the lack of large-scale, real-world, and spatial-temporally aligned event-image datasets. To this end, we propose a real-world (indoor and outdoor) dataset comprising over 30K pairs of images and events under both low and normal illumination conditions. To achieve this, we utilize a robotic arm that traces a consistent non-linear trajectory to curate the dataset with spatial alignment precision under 0.03mm. We then introduce a matching alignment strategy, rendering 90% of our dataset with errors less than 0.01s. Based on the dataset, we propose a novel event-guided LIE approach, called EvLight, towards robust performance in real-world low-light scenes. Specifically, we first design the multiscale holistic fusion branch to extract holistic structural and textural information from both events and images. To ensure robustness against variations in the regional illumination and noise, we then introduce a Signal-to-Noise-Ratio (SNR)-guided regional feature selection to selectively fuse features of images from regions with high SNR and enhance those with low SNR by extracting regional structure information from events. Extensive experiments on our dataset and the synthetic SDSD dataset demonstrate our EvLight significantly surpasses the frame-based methods, e.g., [4] by 1.14 dB and 2.62 dB, respectively.

We present UniBind, a flexible and efficient approach that learns a unified representation space for seven diverse modalities - image, text, audio, point cloud, thermal, video, and event data. Existing works, e.g., ImageBind [13], treat the image as the central modality and build an image-centered representation space; however, the space may be sub-optimal as it leads to an unbalanced representation space among all modalities. Moreover, the category names are directly used to extract text embeddings for the down-stream tasks, making it hardly possible to represent the se-mantics of multi-modal data. The ‘out-of-the-box’ insight of our UniBind is to make the alignment centers modality-agnostic and further learn a unified and balanced repre-sentation space, empowered by the large language mod-els (LLMs). UniBind is superior in its flexible application to all CLIP-style models and delivers remarkable performance boosts. To make this possible, we 1) construct a knowledge base of text with the help of LLMs and multi-modal LLMs; 2) adaptively build LLM-augmented class-wise embedding centers on top of the knowledge base and encoded visual embeddings; 3) align all the embeddings to the LLM-augmented embedding centers via contrastive learning to achieve a unified and balanced representation space. UniBind shows strong zero-shot recognition performance gains over prior arts by an average of 6.36%. Fi-nally, we achieve new state-of-the-art performance, e.g., a 6.75% gain on ImageNet, on the multi-modal fine-tuning setting while reducing 90% of the learnable parameters.

This paper tackles a novel yet challenging problem: how to transfer knowledge from the emerging Segment Anything Model (SAM) - which reveals impressive zero-shot instance segmentation capacity - to learn a compact panoramic semantic segmentation model, i.e., student, without requiring any labeled data. This poses considerable challenges due to SAM's inability to provide semantic labels and the large capacity gap between SAM and the student. To this end, we propose a novel framework, called GoodSAM, that introduces a teacher assistant (TA) to provide semantic information, integrated with SAM to generate ensemble logits to achieve knowledge transfer. Specifically, we propose a Distortion-Aware Rectification (DAR) module that first addresses the distortion problem of panoramic images by imposing prediction-level consistency and boundary enhancement. This subtly enhances TA's prediction capacity on panoramic images. DAR then incorporates a cross-task complementary fusion block to adaptively merge the predictions of SAM and TA to obtain more reliable ensemble logits. Moreover, we introduce a Multi-level Knowledge Adaptation (MKA) module to efficiently transfer the multi-level feature knowledge from TA and ensemble logits to learn a compact student model. Extensive experiments on two benchmarks show that our GoodSAM achieves a remarkable +3.75% mIoU improvement over the state-of-the-art (SOTA) domain adaptation methods, e.g., [41]. Also, our most lightweight model achieves comparable performance to the SOTA methods with only 3.7M parameters.

In this paper, we make the first attempt at achieving the cross-modal (i.e., image-to-events) adaptation for event-based object recognition without accessing any labeled source image data owning to privacy and commercial issues. Tackling this novel problem is non-trivial due to the novelty of event cameras and the distinct modality gap between images and events. In particular, as only the source model is available, a hurdle is how to extract the knowledge from the source model by only using the unlabeled target event data while achieving knowledge transfer. To this end, we propose a novel framework, dubbed Event-Dance for this unsupervised source-free cross-modal adaptation problem. Importantly, inspired by event-to-video reconstruction methods, we propose a reconstruction-based modality bridging (RMB) module, which reconstructs intensity frames from events in a self-supervised manner. This makes it possible to build up the surrogate images to extract the knowledge (i.e., labels) from the source model. We then propose a multi-representation knowledge adaptation (MKA) module that transfers the knowledge to target models learning events with multiple representation types for fully exploring the spatiotemporal information of events. The two modules connecting the source and target models are mutually updated so as to achieve the best performance. Experiments on three benchmark datasets with two adaption settings show that EventDance is on par with prior methods utilizing the source data.

This paper addresses an interesting yet challenging problem-source-free unsupervised domain adaptation (SFUDA) for pinhole-to-panoramic semantic segmentation-given only a pinhole image-trained model (i.e., source) and unlabeled panoramic images (i.e., target). Tackling this problem is nontrivial due to the semantic mismatches, style discrepancies, and inevitable distortion of panoramic images. To this end, we propose a novel method that utilizes Tangent Projection (TP) as it has less distortion and meanwhile slits the equirectangular projection (ERP) with a fixed FoV to mimic the pinhole images. Both projections are shown effective in extracting knowledge from the source model. However, the distinct projection discrepancies between source and target domains impede the direct knowledge transfer; thus, we propose a panoramic prototype adaptation module (PPAM) to integrate panoramic prototypes from the extracted knowledge for adaptation. We then impose the loss constraints on both predictions and prototypes and propose a cross-dual attention module (CDAM) at the feature level to better align the spatial and channel characteristics across the domains and projections. Both knowledge extraction and transfer processes are synchronously updated to reach the best performance. Extensive experiments on the synthetic and real-world benchmarks, including outdoor and indoor scenarios, demonstrate that our method achieves significantly better performance than prior SFUDA methods for pinhole-to-panoramic adaptation.

Event cameras have recently been shown beneficial for practical vision tasks, such as action recognition, thanks to their high temporal resolution, power efficiency, and reduced privacy concerns. However, current research is hindered by 1) the difficulty in processing events because of their prolonged duration and dynamic actions with complex and ambiguous semantics and 2) the redundant action depiction of the event frame representation with fixed stacks. We find language naturally conveys abundant semantic information, rendering it stunningly superior in reducing semantic uncertainty. In light of this, we propose ExACT, a novel approach that, for the first time, tackles event-based action recognition from a cross-modal conceptualizing perspective. Our ExACT brings two technical contributions. Firstly, we propose an adaptive fine-grained event (AFE) representation to adaptively filter out the repeated events for the stationary objects while preserving dynamic ones. This subtly enhances the performance of ExACT without extra computational cost. Then, we propose a conceptual reasoning-based uncertainty estimation module, which simulates the recognition process to enrich the semantic representation. In particular, conceptual reasoning builds the temporal relation based on the action semantics, and uncertainty estimation tackles the semantic uncertainty of actions based on the distributional representation. Experiments show that our ExACT achieves superior recognition accuracy of 94.83%(+2.23%), 90.10%(+37.47%) and 67.24% on PAF, HARDVS and our SeAct datasets respectively.

The popular methods for semi-supervised semantic segmentation mostly adopt a unitary network model using convolutional neural networks (CNNs) and enforce consistency of the model’s predictions over perturbations applied to the inputs or model. However, such a learning paradigm suffers from two critical limitations: a) learning the discriminative features for the unlabeled data; b) learning both global and local information from the whole image. In this paper, we propose a novel Semi-supervised Learning (SSL) approach, called Transformer-CNN Cohort (TCC), that consists of two students with one based on the vision transformer (ViT) and the other based on the CNN. Our method subtly incorporates the multi-level consistency regularization on the predictions and the heterogeneous feature spaces via pseudo-labeling for the unlabeled data. First, as the inputs of the ViT student are image patches, the feature maps extracted encode crucial class-wise statistics. To this end, we propose class-aware feature consistency distillation (CFCD) that first leverages the outputs of each student as the pseudo labels and generates class-aware feature (CF) maps for knowledge transfer between the two students. Second, as the ViT student has more uniform representations for all layers, we propose consistency-aware cross distillation (CCD) to transfer knowledge between the pixel-wise predictions from the cohort. We validate the TCC framework on Cityscapes and Pascal VOC 2012 datasets, which outperforms existing SSL methods by a large margin. Project page: https://vlislab22.github.io/TCC/.

The ability to detect objects in all lighting (i.e., normal-, over-, and under-exposed) conditions is crucial for real-world applications, such as self-driving. Traditional RGB-based detectors often fail under such varying lighting conditions. Therefore, recent works utilize novel event cameras to supplement or guide the RGB modality; however, these methods typically adopt asymmetric network structures that rely predominantly on the RGB modality, resulting in limited robustness for all-day detection. In this paper, we propose EOLO, a novel object detection framework that achieves robust and efficient all-day detection by fusing both RGB and event modalities. Our EOLO framework is built based on a lightweight spiking neural network (SNN) to efficiently leverage the asynchronous property of events. Buttressed by it, we first introduce an Event Temporal Attention (ETA) module to learn the high temporal information from events while preserving crucial edge information. Secondly, as different modalities exhibit varying levels of importance under diverse lighting conditions, we propose a novel Symmetric RGB-Event Fusion (SREF) module to effectively fuse RGB-Event features without relying on a specific modality, thus ensuring a balanced and adaptive fusion for all-day detection. In addition, to compensate for the lack of paired RGB-Event datasets for all-day training and evaluation, we propose an event synthesis approach based on the randomized optical flow that allows for directly generating the event frame from a single exposure image. We further build two new datasets, E-MSCOCO and E-VOC based on the popular benchmarks MSCOCO and PASCAL VOC. Extensive experiments demonstrate that our EOLO outperforms the state-of-the-art detectors, e.g., RENet [1], by a substantial margin (+3.74% mAP50) in all lighting conditions. Our code and datasets will be available at https://vlislab22.github.io/EOLO/.

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 [6], 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 [6], especially in night scenes. Our project homepage: https://vlislab22.github.io/SRFNet.

Nighttime semantic segmentation plays a crucial role in practical applications, such as autonomous driving, where it frequently encounters difficulties caused by inadequate illumination conditions and the absence of well-annotated datasets. Moreover, semantic segmentation models trained on daytime datasets often face difficulties in generalizing effectively to nighttime conditions. Unsupervised domain adaptation (UDA) has shown the potential to address the challenges and achieved remarkable results for nighttime semantic segmentation. However, existing methods still face limitations in 1) their reliance on style transfer or relighting models, which struggle to generalize to complex nighttime environments, and 2) their ignorance of dynamic and small objects like vehicles and poles, which are difficult to be directly learned from other domains. This paper proposes a novel UDA method that refines both label and feature levels for dynamic and small objects for nighttime semantic segmentation. First, we propose a dynamic and small object refinement module to complement the knowledge of dynamic and small objects from the source domain to target the nighttime domain. These dynamic and small objects are normally context-inconsistent in under-exposed conditions. Then, we design a feature prototype alignment module to reduce the domain gap by deploying contrastive learning between features and prototypes of the same class from different domains, while re-weighting the categories of dynamic and small objects. Extensive experiments on three benchmark datasets demonstrate that our method outperforms prior arts by a large margin for nighttime segmentation. Project page: https://rorisis.github.io/DSRNSS/.

Text-to-3D content creation has recently received much attention, especially with the prevalence of 3D Gaussians Splatting (3D GS). 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 (or harder) 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 state-of-the-art training frameworks, e.g., LucidDreamer for semantically consistent text-to-3D generation. The project code is available at https://vlislab22.github.io/DreamInit/.

Event cameras harness advantages such as low latency, high temporal resolution, and high dynamic range (HDR), compared to standard cameras. Due to the distinct imaging paradigm shift, a dominant line of research focuses on event-to-video (E2V) reconstruction to bridge event-based and standard computer vision. However, this task remains challenging due to its inherently ill-posed nature: event cameras only detect the edge and motion information locally. Consequently, the reconstructed videos are often plagued by artifacts and regional blur, primarily caused by the ambiguous semantics of event data. In this paper, we find language naturally conveys abundant semantic information, rendering it stunningly superior in ensuring semantic consistency for E2V reconstruction. Accordingly, we propose a novel framework, called LaSe-E2V, that can achieve semantic-aware high-quality E2V reconstruction from a language-guided perspective, buttressed by the text-conditional diffusion models. However, due to diffusion models' inherent diversity and randomness, it is hardly possible to directly apply them to achieve spatial and temporal consistency for E2V reconstruction. Thus, we first propose an Event-guided Spatiotemporal Attention (ESA) module to condition the event data to the denoising pipeline effectively. We then introduce an event-aware mask loss to ensure temporal coherence and a noise initialization strategy to enhance spatial consistency. Given the absence of event-text-video paired data, we aggregate existing E2V datasets and generate textual descriptions using the tagging models for training and evaluation. Extensive experiments on three datasets covering diverse challenging scenarios (e.g., fast motion, low light) demonstrate the superiority of our method. Demo videos for the results are attached to the project page.

360° images, with a field-of-view (FoV) of $180^{\circ}\times 360^{\circ}$, provide immersive and realistic environments for emerging virtual reality (VR) applications, such as virtual tourism, where users desire to create diverse panoramic scenes from a narrow FoV photo they take from a viewpoint via portable devices. It thus brings us to a technical challenge: 'How to allow the users to freely create diverse and immersive virtual scenes from a narrow FoV image with a specified viewport?' To this end, we propose a transformer-based 360° image outpainting framework called Dream360, which can generate diverse, high-fidelity, and high-resolution panoramas from user-selected viewports, considering the spherical properties of 360° images. Compared with existing methods, e.g., [3], which primarily focus on inputs with rectangular masks and central locations while overlooking the spherical property of 360° images, our Dream360 offers higher outpainting flexibility and fidelity based on the spherical representation. Dream360 comprises two key learning stages: (I) codebook-based panorama outpainting via Spherical-VQGAN (S-VQGAN), and (II) frequency-aware refinement with a novel frequency-aware consistency loss. Specifically, S-VQGAN learns a sphere-specific codebook from spherical harmonic (SH) values, providing a better representation of spherical data distribution for scene modeling. The frequency-aware refinement matches the resolution and further improves the semantic consistency and visual fidelity of the generated results. Our Dream360 achieves significantly lower Frechet Inception Distance (FID) scores and better visual fidelity than existing methods. We also conducted a user study involving 15 participants to interactively evaluate the quality of the generated results in VR, demonstrating the flexibility and superiority of our Dream360 framework.

Eye tracking has shown great promise in many scientific fields and daily applications, ranging from the early detection of mental health disorders to foveated rendering in virtual reality (VR). These applications all call for a robust system for high-frequency near-eye movement sensing and analysis in high precision, which cannot be guaranteed by the existing eye tracking solutions with CCD/CMOS cameras. To bridge the gap, in this paper, we propose Swift-Eye, an offline precise and robust pupil estimation and tracking framework to support high-frequency near-eye movement analysis, especially when the pupil region is partially occluded. Swift-Eye is built upon the emerging event cameras to capture the high-speed movement of eyes in high temporal resolution. Then, a series of bespoke components are designed to generate high-quality near-eye movement video at a high frame rate over kilohertz and deal with the occlusion over the pupil caused by involuntary eye blinks. According to our extensive evaluations on EV-Eye, a large-scale public dataset for eye tracking using event cameras, Swift-Eye shows high robustness against significant occlusion. It can improve the IoU and F1-score of the pupil estimation by 20% and 12.5% respectively, compared with the second-best competing approach, when over 80% of the pupil region is occluded by the eyelid. Lastly, it provides continuous and smooth traces of pupils in extremely high temporal resolution and can support high-frequency eye movement analysis and a number of potential applications, such as mental health diagnosis, behaviour-brain association, etc. The implementation details and source codes can be found at https://github.com/ztysdu/Swift-Eye.

Recently, the self-supervised pre-training paradigm has shown great potential in leveraging large-scale unlabeled data to improve downstream task performance. However, increasing the scale of unlabeled pre-training data in real-world scenarios requires prohibitive computational costs and faces the challenge of uncurated samples. To address these issues, we build a task-specific self-supervised pre-training framework from a data selection perspective based on a simple hypothesis that pre-training on the unlabeled samples with similar distribution to the target task can bring substantial performance gains. Buttressed by the hypothesis, we propose the first yet novel framework for Scalable and Efficient visual Pre-Training (SEPT) by introducing a retrieval pipeline for data selection. SEPT first leverage a self-supervised pre-trained model to extract the features of the entire unlabeled dataset for retrieval pipeline initialization. Then, for a specific target task, SEPT retrievals the most similar samples from the unlabeled dataset based on feature similarity for each target instance for pre-training. Finally, SEPT pre-trains the target model with the selected unlabeled samples in a self-supervised manner for target data finetuning. By decoupling the scale of pre-training and available upstream data for a target task, SEPT achieves high scalability of the upstream dataset and high efficiency of pre-training, resulting in high model architecture flexibility. Results on various downstream tasks demonstrate that SEPT can achieve competitive or even better performance compared with ImageNet pre-training while reducing the size of training samples by one magnitude without resorting to any extra annotations.

Depth estimation from a monocular 360° image is a burgeoning problem owing to its holistic sensing of a scene. Recently, some methods, e.g., OmniFusion, have applied the tangent projection (TP) to represent a 360° image and predicted depth values via patch-wise regressions, which are merged to get a depth map with equirectangular projection (ERP) format. However, these methods suffer from 1) non-trivial process of merging plenty of patches; 2) capturing less holistic-with-regional contextual information by directly regressing the depth value of each pixel. In this paper, we propose a novel framework, HRDFuse, that subtly combines the potential of convolutional neural networks (CNNs) and transformers by collaboratively learning the holistic contextual information from the ERP and the regional structural information from the TP. Firstly, we propose a spatial feature alignment (SFA) module that learns feature similarities between the TP and ERP to aggregate the TP features into a complete ERP feature map in a pixelwise manner. Secondly, we propose a collaborative depth distribution classification (CDDC) module that learns the holistic-with-regional histograms capturing the ERP and TP depth distributions. As such, the final depth values can be predicted as a linear combination of histogram bin centers. Lastly, we adaptively combine the depth predictions from ERP and TP to obtain the final depth map. Extensive experiments show that our method predicts more smooth and accurate depth results while achieving favorably better results than the SOTA methods.

Event cameras sense the intensity changes asynchronously and produce event streams with high dynamic range and low latency. This has inspired research endeavors utilizing events to guide the challenging video super-resolution (VSR) task. In this paper, we make the first attempt to address a novel problem of achieving VSR at random scales by taking advantages of the high temporal resolution property of events. This is hampered by the difficulties of representing the spatial-temporal information of events when guiding VSR. To this end, we propose a novel framework that incorporates the spatial-temporal interpolation of events to VSR in a unified framework. Our key idea is to learn implicit neural representations from queried spatial-temporal coordinates and features from both RGB frames and events. Our method contains three parts. Specifically, the Spatial-Temporal Fusion (STF) module first learns the 3D features from events and RGB frames. Then, the Temporal Filter (TF) module unlocks more explicit motion information from the events near the queried timestamp and generates the 2D features. Lastly, the Spatial-Temporal Implicit Representation (STIR) module recovers the SR frame in arbitrary resolutions from the outputs of these two modules. In addition, we collect a real-world dataset with spatially aligned events and RGB frames. Extensive experiments show that our method significantly surpasses the prior-arts and achieves VSR with random scales, e.g., 6.5. Code and dataset are available at https://vlis2022.github.io/cvpr23/egvsr.

The ability of scene understanding has sparked active research for panoramic image semantic segmentation. However, the performance is hampered by distortion of the equirectangular projection (ERP) and a lack of pixel-wise annotations. For this reason, some works treat the ERP and pinhole images equally and transfer knowledge from the pinhole to ERP images via unsupervised domain adaptation (UDA). However, they fail to handle the domain gaps caused by: 1) the inherent differences between camera sensors and captured scenes; 2) the distinct image formats (e.g., ERP and pinhole images). In this paper, we propose a novel yet flexible dual-path UDA framework, DPPASS, taking ERP and tangent projection (TP) images as inputs. To reduce the domain gaps, we propose cross-projection and intra-projection training. The cross-projection training includes tangent-wise feature contrastive training and prediction consistency training. That is, the former formulates the features with the same projection locations as positive examples and vice versa, for the models' awareness of distortion, while the latter ensures the consistency of cross-model predictions between the ERP and TP. Moreover, adversarial intra-projection training is proposed to reduce the inherent gap, between the features of the pinhole images and those of the ERP and TP images, respectively. Importantly, the TP path can be freely removed after training, leading to no additional inference cost. Extensive experiments on two benchmarks show that our DPPASS achieves + 1.06% mIoU increment than the state-of-the-art approaches. https://vlis2022.github.io/cvpr23/DPPASS

Endeavors have been recently made to leverage the vision transformer (ViT) for the challenging unsupervised domain adaptation (UDA) task. They typically adopt the cross-attention in ViT for direct domain alignment. However, as the performance of cross-attention highly relies on the quality of pseudo labels for targeted samples, it becomes less effective when the domain gap becomes large. We solve this problem from a game theory's perspective with the proposed model dubbed as PMTrans, which bridges source and target domains with an intermediate domain. Specifically, we propose a novel ViT-based module called PatchMix that effectively builds up the intermediate domain, i.e., probability distribution, by learning to sample patches from both domains based on the game-theoretical models. This way, it learns to mix the patches from the source and target domains to maximize the cross entropy (CE), while exploiting two semi-supervised mixup losses in the feature and label spaces to minimize it. As such, we interpret the process of UDA as a min-max CE game with three players, including the feature extractor, classifier, and PatchMix, to find the Nash Equilibria. Moreover, we leverage attention maps from ViT to re-weight the label of each patch by its importance, making it possible to obtain more domain-discriminative feature representations. We conduct extensive experiments on four benchmark datasets, and the results show that PMTrans significantly surpasses the ViT-based and CNN-based SoTA methods by +3.6% on Office-Home, +1.4% on Office-31, and +17.7% on DomainNet, respectively. https://vlis2022.github.io/cvpr23/PMTrans

The explicit neural radiance field (NeRF) has gained considerable interest for its efficient training and fast inference capabilities, making it a promising direction such as virtual reality and gaming. In particular, PlenOctree (POT) [43], an explicit hierarchical multi-scale octree representation, has emerged as a structural and influential framework. However, POT’s fixed structure for direct optimization is sub-optimal as the scene complexity evolves continuously with updates to cached color and density, necessitating refining the sampling distribution to capture signal complexity accordingly. To address this issue, we propose the dynamic PlenOctree (DOT), which adaptively refines the sample distribution to adjust to changing scene complexity. Specifically, DOT proposes a concise yet novel hierarchical feature fusion strategy during the iterative rendering process. Firstly, it identifies the regions of interest through training signals to ensure adaptive and efficient refinement. Next, rather than directly filtering out valueless nodes, DOT introduces the sampling and pruning operations for octrees to aggregate features, enabling rapid parameter learning. Compared with POT, our DOT outperforms it by enhancing visual quality, reducing over 55.15/68.84% parameters, and providing 1.7/1.9 times FPS for NeRF-synthetic and Tanks & Temples, respectively.

Omnidirectional images (ODIs) have become increasingly popular, as their large field-of-view (FoV) can offer viewers the chance to freely choose the view directions in immersive environments such as virtual reality. The Möbius transformation is typically employed to further provide the opportunity for movement and zoom on ODIs, but applying it to the image level often results in blurry effect and aliasing problem. In this paper, we propose a novel deep learning-based approach, called OmniZoomer, to incorporate the Möbius transformation into the network for movement and zoom on ODIs. By learning various transformed feature maps under different conditions, the network is enhanced to handle the increasing edge curvatures, which alleviates the blurry effect. Moreover, to address the aliasing problem, we propose two key components. Firstly, to compensate for the lack of pixels for describing curves, we enhance the feature maps in the high-resolution (HR) space and calculate the transformed index map with a spatial index generation module. Secondly, considering that ODIs are inherently represented in the spherical space, we propose a spherical resampling module that combines the index map and HR feature maps to transform the feature maps for better spherical correlation. The transformed feature maps are decoded to output a zoomed ODI. Experiments show that our method can produce HR and high-quality ODIs with the flexibility to move and zoom in to the object of interest. Project page is available at http: //vlislab22.github.io/OmniZoomer/.

Deep Neural Networks (DNNs) for 3D point cloud recognition are vulnerable to adversarial examples, threatening their practical deployment. Despite the many research endeavors have been made to tackle this issue in recent years, the diversity of adversarial examples on 3D point clouds makes them more challenging to defend against than those on 2D images. For examples, attackers can generate adversarial examples by adding, shifting, or removing points. Consequently, existing defense strategies are hard to counter unseen point cloud adversarial examples. In this paper, we first establish a comprehensive, and rigorous point cloud adversarial robustness benchmark to evaluate adversarial robustness, which can provide a detailed understanding of the effects of the defense and attack methods. We then collect existing defense tricks in point cloud adversarial defenses and then perform extensive and systematic experiments to identify an effective combination of these tricks. Furthermore, we propose a hybrid training augmentation methods that consider various types of point cloud adversarial examples to adversarial training, significantly improving the adversarial robustness. By combining these tricks, we construct a more robust defense framework achieving an average accuracy of 83.45% against various attacks, demonstrating its capability to enabling robust learners. Our codebase are open-sourced on: https://github.com/qiufan319/benchmark_pc_attack.git.

Endeavors have been recently made to transfer knowledge from the labeled pinhole image domain to the unlabeled panoramic image domain via Unsupervised Domain Adaptation (UDA). The aim is to tackle the domain gaps caused by the style disparities and distortion problem from the non-uniformly distributed pixels of equirectangular projection (ERP). Previous works typically focus on transferring knowledge based on geometric priors with specially designed multi-branch network architectures. As a result, considerable computational costs are induced, and meanwhile, their generalization abilities are profoundly hindered by the variation of distortion among pixels. In this paper, we find that the pixels’ neighborhood regions of the ERP indeed introduce less distortion. Intuitively, we propose a novel UDA framework that can effectively address the distortion problems for panoramic semantic segmentation. In comparison, our method is simpler, easier to implement, and more computationally efficient. Specifically, we propose distortion-aware attention (DA) capturing the neighboring pixel distribution without using any geometric constraints. Moreover, we propose a class-wise feature aggregation (CFA) module to iteratively update the feature representations with a memory bank. As such, the feature similarity between two domains can be consistently optimized. Extensive experiments show that our method achieves new state-of-the-art performance while remarkably reducing 80% parameters.

In this paper, we strive to answer the question ‘how to collaboratively learn convolutional neural network (CNN)-based and vision transformer (ViT)-based models by selecting and exchanging the reliable knowledge between them for semantic segmentation?’ Accordingly, we propose an online knowledge distillation (KD) framework that can simultaneously learn compact yet effective CNN-based and ViT-based models with two key technical breakthroughs to take full advantage of CNNs and ViT while compensating their limitations. Firstly, we propose heterogeneous feature distillation (HFD) to improve students’ consistency in low-layer feature space by mimicking heterogeneous features between CNNs and ViT. Secondly, to facilitate the two students to learn reliable knowledge from each other, we propose bidirectional selective distillation (BSD) that can dynamically transfer selective knowledge. This is achieved by 1) region-wise BSD determining the directions of knowledge transferred between the corresponding regions in the feature space and 2) pixel-wise BSD discerning which of the prediction knowledge to be transferred in the logit space. Extensive experiments on three benchmark datasets demonstrate that our proposed framework outperforms the state-of-the-art online distillation methods by a large margin, and shows its efficacy in learning collaboratively between ViT-based and CNN-based models.

The$360^{\circ}$imaging has recently gained much attention; however, its angular resolution is relatively lower than that of a narrow field-of-view (FOV) perspective image as it is captured using a fisheye lens with the same sensor size. Therefore, it is beneficial to super-resolve a$360^{\circ}$image. Several attempts have been made, but mostly considered equirectangular projection (ERP) as one of the ways for$360^{\circ}$image representation despite the latitude-dependent distortions. In that case, as the output high-resolution (HR) image is always in the same ERP format as the low-resolution (LR) input, additional information loss may occur when transforming the HR image to other projection types. In this paper, we propose SphereSR, a novel framework to generate a continuous spherical image representation from an LR$360^{\circ}$image, with the goal of predicting the RGB values at given spherical coordinates for super-resolution with an arbitrary$360^{\circ}$image projection. Specifically, first we propose a feature extraction module that represents the spherical data based on an icosahedron and that efficiently extracts features on the spherical surface. We then propose a spherical local implicit image function (SLIIF) to predict RGB values at the spherical coordinates. As such, SphereSR flexibly reconstructs an HR image given an arbitrary projection type. Experiments on various benchmark datasets show that the proposed method significantly surpasses existing methods in terms of performance.

Event cameras sense brightness changes in each pixel and yield asynchronous event streams instead of producing intensity images. They have distinct advantages over conventional cameras, such as a high dynamic range (HDR) and no motion blur. To take advantage of event cameras with existing image-based algorithms, a few methods have been proposed to reconstruct images from event streams. However, the output images have a low resolution (LR) and are unrealistic. Low-quality outputs stem from broader applications of event cameras, where high-quality and high-resolution (HR) images are needed. In this work, we consider the problem of reconstructing and super-resolving images from LR events when no ground truth (GT) HR images and degradation models are available. We propose a novel end-to-end joint framework for single image reconstruction and super-resolution from LR event data. Our method is primarily unsupervised to handle the absence of real inputs from GT and deploys adversarial learning. To train our framework, we constructed an open dataset, including simulated events and real-world images. The use of the dataset boosts the network performance, and the network architectures and various loss functions in each phase help improve the quality of the resulting image. Various experiments showed that our method surpasses the state-of-the-art LR image reconstruction methods for real-world and synthetic datasets. The experiments for super-resolution (SR) image reconstruction also substantiate the effectiveness of the proposed method. We further extended our method to more challenging problems of HDR, sharp image reconstruction, and color events. In addition, we demonstrate that the reconstruction and super-resolution results serve as intermediate representations of events for high-level tasks, such as semantic segmentation, object recognition, and detection. We further examined how events affect the outputs of the three phases and analyze our method's efficacy through an ablation study.

Deep neural models, in recent years, have been successful in almost every field, even solving the most complex problem statements. However, these models are huge in size with millions (and even billions) of parameters, demanding heavy computation power and failing to be deployed on edge devices. Besides, the performance boost is highly dependent on redundant labeled data. To achieve faster speeds and to handle the problems caused by the lack of labeled data, knowledge distillation (KD) has been proposed to transfer information learned from one model to another. KD is often characterized by the so-called 'Student-Teacher' (S-T) learning framework and has been broadly applied in model compression and knowledge transfer. This paper is about KD and S-T learning, which are being actively studied in recent years. First, we aim to provide explanations of what KD is and how/why it works. Then, we provide a comprehensive survey on the recent progress of KD methods together with S-T frameworks typically used for vision tasks. In general, we investigate some fundamental questions that have been driving this research area and thoroughly generalize the research progress and technical details. Additionally, we systematically analyze the research status of KD in vision applications. Finally, we discuss the potentials and open challenges of existing methods and prospect the future directions of KD and S-T learning.

High dynamic range (HDR) imaging is a technique that allows an extensive dynamic range of exposures, which is important in image processing, computer graphics, and computer vision. In recent years, there has been a significant advancement in HDR imaging using deep learning (DL). This study conducts a comprehensive and insightful survey and analysis of recent developments in deep HDR imaging methodologies. We hierarchically and structurally group existing deep HDR imaging methods into five categories based on (1) number/domain of input exposures, (2) number of learning tasks, (3) novel sensor data, (4) novel learning strategies, and (5) applications. Importantly, we provide a constructive discussion on each category regarding its potential and challenges. Moreover, we review some crucial aspects of deep HDR imaging, such as datasets and evaluation metrics. Finally, we highlight some open problems and point out future research directions.

Event cameras sense per-pixel intensity changes and produce asynchronous event streams with high dynamic range and less motion blur, showing advantages over the conventional cameras. A hurdle of training event-based models is the lack of large qualitative labeled data. Prior works learning end-tasks mostly rely on labeled or pseudo-labeled datasets obtained from the active pixel sensor (APS) frames; however, such datasets’ quality is far from rivaling those based on the canonical images. In this paper, we propose a novel approach, called EvDistill, to learn a student network on the unlabeled and unpaired event data (target modality) via knowledge distillation (KD) from a teacher network trained with large-scale, labeled image data (source modality). To enable KD across the unpaired modalities, we first propose a bidirectional modality reconstruction (BMR) module to bridge both modalities and simultaneously exploit them to distill knowledge via the crafted pairs, causing no extra computation in the inference. The BMR is improved by the end-tasks and KD losses in an end-to-end manner. Second, we leverage the structural similarities of both modalities and adapt the knowledge by matching their distributions. Moreover, as most prior feature KD methods are uni-modality and less applicable to our problem, we propose an affinity graph KD loss to boost the distillation. Our extensive experiments on semantic segmentation and object recognition demonstrate that EvDistill achieves significantly better results than the prior works and KD with only events and APS frames.

Event cameras are novel sensors that perceive the perpixel intensity changes and output asynchronous event streams with high dynamic range and less motion blur. It has been shown that events alone can be used for end-task learning, e.g., semantic segmentation, based on encoder-decoder-like networks. However, as events are sparse and mostly reflect edge information, it is difficult to recover original details merely relying on the decoder. Moreover, most methods resort to the pixel-wise loss alone for supervision, which might be insufficient to fully exploit the visual details from sparse events, thus leading to less optimal performance. In this paper, we propose a simple yet flexible two-stream framework named Dual Transfer Learning (DTL) to effectively enhance the performance on the end-tasks without adding extra inference cost. The proposed approach consists of three parts: event to end-task learning (EEL) branch, event to image translation (EIT) branch, and transfer learning (TL) module that simultaneously explores the feature-level affinity information and pixel-level knowledge from the EIT branch to improve the EEL branch. This simple yet novel method leads to strong representation learning from events and is evidenced by the significant performance boost on the end-tasks such as semantic segmentation and depth estimation.

Recent advances in deep neural networks (DNNs) have facilitated high-end applications, including holistic scene understanding (HSU), in which many tasks run in parallel with the same visual input. Following this trend, various methods have been proposed to use DNNs to perform multiple vision tasks. However, these methods are task-specific and less effective when considering multiple HSU tasks. End-to-end demonstrations of adversarial examples, which generate one-to-many heterogeneous adversarial examples in parallel from the same input, are scarce. Additionally, one-to-many mapping of adversarial examples for HSU usually requires joint representation learning and flexible constraints on magnitude, which can render the prevalent attack methods ineffective. In this paper, we propose PSAT-GAN, an end-to-end framework that follows the pipeline of HSU. It is based on a mixture of generative models and an adversarial classifier that employs partial weight sharing to learn a one-to-many mapping of adversarial examples in parallel, each of which is effective for its corresponding task in HSU attacks. PSAT-GAN is further enhanced by applying novel adversarial and soft-constraint losses to generate effective perturbations and avoid studying transferability. Experimental results indicate that our method is efficient in generating both universal and image-dependent adversarial examples to fool HSU tasks under either targeted or non-targeted settings.

Event cameras sense intensity changes and have many advantages over conventional cameras. To take advantage of event cameras, some methods have been proposed to reconstruct intensity images from event streams. However, the outputs are still in low resolution (LR), noisy, and unrealistic. The low-quality outputs stem broader applications of event cameras, where high spatial resolution (HR) is needed as well as high temporal resolution, dynamic range, and no motion blur. We consider the problem of reconstructing and super-resolving intensity images from pure events, when no ground truth (GT) HR images and down-sampling kernels are available. To tackle the challenges, we propose a novel end-to-end pipeline that reconstructs LR images from event streams, enhances the image qualities and upsamples the enhanced images, called EventSR. For the absence of real GT images, our method is primarily unsupervised, deploying adversarial learning. To train EventSR, we create an open dataset including both real-world and simulated scenes. The use of both datasets boosts up the network performance, and the network architectures and various loss functions in each phase help improve the image qualities. The whole pipeline is trained in three phases. While each phase is mainly for one of the three tasks, the networks in earlier phases are fine-tuned by respective loss functions in an end-to-end manner. Experimental results show that EventSR generates high-quality SR images from events for both simulated and real-world data.

Event cameras have a lot of advantages over traditional cameras, such as low latency, high temporal resolution, and high dynamic range. However, since the outputs of event cameras are the sequences of asynchronous events over time rather than actual intensity images, existing algorithms could not be directly applied. Therefore, it is demanding to generate intensity images from events for other tasks. In this paper, we unlock the potential of event camera-based conditional generative adversarial networks to create images/videos from an adjustable portion of the event data stream. The stacks of space-time coordinates of events are used as inputs and the network is trained to reproduce images based on the spatio-temporal intensity changes. The usefulness of event cameras to generate high dynamic range (HDR) images even in extreme illumination conditions and also non blurred images under rapid motion is also shown. In addition, the possibility of generating very high frame rate videos is demonstrated, theoretically up to 1 million frames per second(FPS) since the temporal resolution of event cameras is about 1 microsecond. Proposed methods are evaluated by comparing the results with the intensity images captured on the same pixel grid-line of events using online available real datasets and synthetic datasets produced by the event camera simulator.

Abstract The development of ocean floating wind power has been burgeoning in recent years because of its low cost and high efficiency. To facilitate effectiveness, 3D visualization using virtual reality and augmented reality technologies has been applied to many operating systems. However, most of the existing 3D motion visualizations are “pseudo” visualization, and there are a few realistic visualization systems that base the motion of ocean floating wind power on simulation and experiment results. Therefore, in this paper, we conducted research related to the design for a realistic motion visualization system based on numerical simulation data using a commercial game engine (Unity 3D). In our system, the six‐degree‐of‐freedom motion (Surge, Sway, Heave, Roll, Pitch, and Yaw) is simulated and visualized based on numerical analysis results of two hydrodynamics simulation softwares, which can illuminate the nuance between simulation results and experiment results and give us a “real‐time” visual experience about motion in each direction. Meanwhile, comprehensive sea environment conditions, such as wind, rain, water, sound, and cloudiness, are also visualized in Unity 3D.