Search

Detecting AI-generated images has become an extraordinarily difficult challenge as new generative architectures emerge on a daily basis with more and more capabilities and unprecedented realism. New versions of many commercial tools, such as DALLE, Midjourney, and Stable Diffusion, have been released recently, and it is impractical to continually update and retrain supervised forensic detectors to handle such a large variety of models. To address this challenge, we propose a zero-shot entropy-based detector (ZED) that neither needs AI-generated training data nor relies on knowledge of generative architectures to artificially synthesize their artifacts. Inspired by recent works on machine-generated text detection, our idea is to measure how surprising the image under analysis is compared to a model of real images. To this end, we rely on a lossless image encoder that estimates the probability distribution of each pixel given its context. To ensure computational efficiency, the encoder has a multi-resolution architecture and contexts comprise mostly pixels of the lower-resolution version of the image.Since only real images are needed to learn the model, the detector is independent of generator architectures and synthetic training data. Using a single discriminative feature, the proposed detector achieves state-of-the-art performance. On a wide variety of generative models it achieves an average improvement of more than 3% over the SoTA in terms of accuracy. Code is available at https://grip-unina.github.io/ZED/.

High frame rate and accurate depth estimation plays an important role in several tasks crucial to robotics and automotive perception. To date, this can be achieved through ToF and LiDAR devices for indoor and outdoor applications, respectively. However, their applicability is limited by low frame rate, energy consumption, and spatial sparsity. Depth on Demand (DoD) allows for accurate temporal and spatial depth densification achieved by exploiting a high frame rate RGB sensor coupled with a potentially lower frame rate and sparse active depth sensor. Our proposal jointly enables lower energy consumption and denser shape reconstruction, by significantly reducing the streaming requirements on the depth sensor thanks to its three core stages: i) multi-modal encoding, ii) iterative multi-modal integration, and iii) depth decoding. We present extended evidence assessing the effectiveness of DoD on indoor and outdoor video datasets, covering both environment scanning and automotive perception use cases., Comment: Accepted for publication at the European Conference on Computer Vision (ECCV) 2024

3D Gaussian Splatting (GS) significantly struggles to accurately represent the underlying 3D scene geometry, resulting in inaccuracies and floating artifacts when rendering depth maps. In this paper, we address this limitation, undertaking a comprehensive analysis of the integration of depth priors throughout the optimization process of Gaussian primitives, and present a novel strategy for this purpose. This latter dynamically exploits depth cues from a readily available stereo network, processing virtual stereo pairs rendered by the GS model itself during training and achieving consistent self-improvement of the scene representation. Experimental results on three popular datasets, breaking ground as the first to assess depth accuracy for these models, validate our findings., Comment: BMVC 2024. Project page: https://kuis-ai.github.io/StereoGS/

SPIDER (Source for the Production of Ions of Deuterium Extracted from Rf plasma) is a full-scale prototype of the ITER NBI source. It is based on the concept of inductive coupling between radio-frequency current drive and plasma. Present three-dimensional (3D) electromagnetic calculations of stationary RF fields in SPIDER permit an evaluation of the power dissipation in its main constituents. Taking experimental plasma parameters as input, we concentrate on the power dissipation in the copper-made Faraday shield lateral wall (FSLW) of the source for discharges with and without a static magnetic filter field. In agreement with our previous results and a first comparison with calorimetry data from the FSLW cooling circuit, the FSLW cylinder alone absorbs around 50\% of the available power for the studied plasma parameters. A hypothesized improvement of transport confinement may increase significantly the efficiency., Comment: 14 pages, 8 figures

One of the main statistical features of near-neutral atmospheric boundary layer (ABL) turbulence is the positive vertical velocity skewness $Sk_w$ above the roughness sublayer or the buffer region in smooth-walls. The $Sk_w$ variations are receiving renewed interest in many climate-related parameterizations of the ABL given their significance to cloud formation and to testing sub-grid schemes for Large Eddy Simulations (LES). The vertical variations of $Sk_w$ are explored here using high Reynolds number wind tunnel and flume experiments collected above smooth, rough, and permeable-walls in the absence of buoyancy and Coriolis effects. These laboratory experiments form a necessary starting point to probe the canonical structure of $Sk_w$ as they deal with a key limiting case (i.e. near-neutral conditions) that has received much less attention compared to its convective counterpart in atmospheric turbulence studies. Diagnostic models based on cumulant expansions, realizability constraints, and the now-popular constant mass flux approach routinely employed in the convective boundary layer as well as prognostic models based on third-order budgets are used to explain variations in $Sk_w$ for the idealized laboratory conditions. The failure of flux-gradient relations to model $Sk_w$ from the gradients of the vertical velocity variance $\sigma_w^2$ are explained and corrections based on models of energy transport offered. Novel links between the diagnostic and prognostic models are also featured, especially for the inertial term in the third order budget of the vertical velocity fluctuation. The co-spectral properties of $w'/\sigma_w$ versus $w'^2/\sigma_w^2$ are also presented for the first time to assess the dominant scales governing $Sk_w$ in the inner and outer layers, where $w'$ is the fluctuating vertical velocity and $\sigma_w$ is the vertical velocity standard deviation.

Event stereo matching is an emerging technique to estimate depth from neuromorphic cameras; however, events are unlikely to trigger in the absence of motion or the presence of large, untextured regions, making the correspondence problem extremely challenging. Purposely, we propose integrating a stereo event camera with a fixed-frequency active sensor -- e.g., a LiDAR -- collecting sparse depth measurements, overcoming the aforementioned limitations. Such depth hints are used by hallucinating -- i.e., inserting fictitious events -- the stacks or raw input streams, compensating for the lack of information in the absence of brightness changes. Our techniques are general, can be adapted to any structured representation to stack events and outperform state-of-the-art fusion methods applied to event-based stereo., Comment: ECCV 2024. Code: https://github.com/bartn8/eventvppstereo/ - Project Page: https://eventvppstereo.github.io/

Homogeneous matroids are characterized by the property that strength equals fractional arboricity, and arise in the study of base modulus [22]. For graphic matroids, Cunningham [9] provided efficient algorithms for calculating graph strength, and also for determining minimum cost reinforcement to achieve a desired strength. This paper extends this latter problem by focusing on two optimal strategies for transforming a matroid into a homogeneous one, by either increasing or decreasing element weights. As an application to graphs, we give algorithms to solve this problem in the context of spanning trees.

There are numerous randomized algorithms to generate spanning trees in a given ambient graph; several target the uniform distribution on trees (UST), while in practice the fastest and most frequently used draw random weights on the edges and then employ a greedy algorithm to choose the minimum-weight spanning tree (MST). Though MST is a workhorse in applications, the mathematical properties of random MST are far less explored than those of UST. In this paper we develop tools for the quantitative study of random MST. We consider the standard case that the weights are drawn i.i.d. from a single distribution on the real numbers, as well as successive generalizations that lead to \emph{product measures}, where the weights are independently drawn from arbitrary distributions.

In recent years, many forensic detectors have been proposed to detect AI-generated images and prevent their use for malicious purposes. Convolutional neural networks (CNNs) have long been the dominant architecture in this field and have been the subject of intense study. However, recently proposed Transformer-based detectors have been shown to match or even outperform CNN-based detectors, especially in terms of generalization. In this paper, we study the adversarial robustness of AI-generated image detectors, focusing on Contrastive Language-Image Pretraining (CLIP)-based methods that rely on Visual Transformer (ViT) backbones and comparing their performance with CNN-based methods. We study the robustness to different adversarial attacks under a variety of conditions and analyze both numerical results and frequency-domain patterns. CLIP-based detectors are found to be vulnerable to white-box attacks just like CNN-based detectors. However, attacks do not easily transfer between CNN-based and CLIP-based methods. This is also confirmed by the different distribution of the adversarial noise patterns in the frequency domain. Overall, this analysis provides new insights into the properties of forensic detectors that can help to develop more effective strategies.

We present a novel approach designed to address the complexities posed by challenging, out-of-distribution data in the single-image depth estimation task. Starting with images that facilitate depth prediction due to the absence of unfavorable factors, we systematically generate new, user-defined scenes with a comprehensive set of challenges and associated depth information. This is achieved by leveraging cutting-edge text-to-image diffusion models with depth-aware control, known for synthesizing high-quality image content from textual prompts while preserving the coherence of 3D structure between generated and source imagery. Subsequent fine-tuning of any monocular depth network is carried out through a self-distillation protocol that takes into account images generated using our strategy and its own depth predictions on simple, unchallenging scenes. Experiments on benchmarks tailored for our purposes demonstrate the effectiveness and versatility of our proposal., Comment: ECCV 2024. Code: https://github.com/fabiotosi92/Diffusion4RobustDepth Project page: https://diffusion4robustdepth.github.io/

Stereo matching is close to hitting a half-century of history, yet witnessed a rapid evolution in the last decade thanks to deep learning. While previous surveys in the late 2010s covered the first stage of this revolution, the last five years of research brought further ground-breaking advancements to the field. This paper aims to fill this gap in a two-fold manner: first, we offer an in-depth examination of the latest developments in deep stereo matching, focusing on the pioneering architectural designs and groundbreaking paradigms that have redefined the field in the 2020s; second, we present a thorough analysis of the critical challenges that have emerged alongside these advances, providing a comprehensive taxonomy of these issues and exploring the state-of-the-art techniques proposed to address them. By reviewing both the architectural innovations and the key challenges, we offer a holistic view of deep stereo matching and highlight the specific areas that require further investigation. To accompany this survey, we maintain a regularly updated project page that catalogs papers on deep stereo matching in our Awesome-Deep-Stereo-Matching (https://github.com/fabiotosi92/Awesome-Deep-Stereo-Matching) repository., Comment: Extended version of CVPR 2024 Tutorial "Deep Stereo Matching in the Twenties" (https://sites.google.com/view/stereo-twenties)

We provide empirical evidence that within the inertial sub layer of adiabatic turbulent flows over smooth walls, the skewness of the vertical velocity component $Sk_w$ displays universal behaviour, being constant and constrained within the range $Sk_w \approx 0.1-0.16$, regardless of flow configuration and Reynolds number. A theoretical model is proposed to explain the observed behaviour, including the observed range of variations of $Sk_w$. The model clarifies why $Sk_w$ cannot be predicted from down-gradient closure approximations routinely employed in meteorological and climate models whereby $Sk_w$ impacts cloud formation and dispersion processes. The model also offers an alternative and implementable approach.

Hyperspectral pansharpening consists of fusing a high-resolution panchromatic band and a low-resolution hyperspectral image to obtain a new image with high resolution in both the spatial and spectral domains. These remote sensing products are valuable for a wide range of applications, driving ever growing research efforts. Nonetheless, results still do not meet application demands. In part, this comes from the technical complexity of the task: compared to multispectral pansharpening, many more bands are involved, in a spectral range only partially covered by the panchromatic component and with overwhelming noise. However, another major limiting factor is the absence of a comprehensive framework for the rapid development and accurate evaluation of new methods. This paper attempts to address this issue. We started by designing a dataset large and diverse enough to allow reliable training (for data-driven methods) and testing of new methods. Then, we selected a set of state-of-the-art methods, following different approaches, characterized by promising performance, and reimplemented them in a single PyTorch framework. Finally, we carried out a critical comparative analysis of all methods, using the most accredited quality indicators. The analysis highlights the main limitations of current solutions in terms of spectral/spatial quality and computational efficiency, and suggests promising research directions. To ensure full reproducibility of the results and support future research, the framework (including codes, evaluation procedures and links to the dataset) is shared on https://github.com/matciotola/hyperspectral_pansharpening_toolbox, as a single Python-based reference benchmark toolbox.

This paper presents a novel general-purpose stereo and depth data fusion paradigm that mimics the active stereo principle by replacing the unreliable physical pattern projector with a depth sensor. It works by projecting virtual patterns consistent with the scene geometry onto the left and right images acquired by a conventional stereo camera, using the sparse hints obtained from a depth sensor, to facilitate the visual correspondence. Purposely, any depth sensing device can be seamlessly plugged into our framework, enabling the deployment of a virtual active stereo setup in any possible environment and overcoming the severe limitations of physical pattern projection, such as the limited working range and environmental conditions. Exhaustive experiments on indoor and outdoor datasets featuring both long and close range, including those providing raw, unfiltered depth hints from off-the-shelf depth sensors, highlight the effectiveness of our approach in notably boosting the robustness and accuracy of algorithms and deep stereo without any code modification and even without re-training. Additionally, we assess the performance of our strategy on active stereo evaluation datasets with conventional pattern projection. Indeed, in all these scenarios, our virtual pattern projection paradigm achieves state-of-the-art performance. The source code is available at: https://github.com/bartn8/vppstereo., Comment: extended version of ICCV 2023: "Active Stereo Without Pattern Projector"

This work addresses the challenge of video depth estimation, which expects not only per-frame accuracy but, more importantly, cross-frame consistency. Instead of directly developing a depth estimator from scratch, we reformulate the prediction task into a conditional generation problem. This allows us to leverage the prior knowledge embedded in existing video generation models, thereby reducing learning difficulty and enhancing generalizability. Concretely, we study how to tame the public Stable Video Diffusion (SVD) to predict reliable depth from input videos using a mixture of image depth and video depth datasets. We empirically confirm that a procedural training strategy -- first optimizing the spatial layers of SVD and then optimizing the temporal layers while keeping the spatial layers frozen -- yields the best results in terms of both spatial accuracy and temporal consistency. We further examine the sliding window strategy for inference on arbitrarily long videos. Our observations indicate a trade-off between efficiency and performance, with a one-frame overlap already producing favorable results. Extensive experimental results demonstrate the superiority of our approach, termed ChronoDepth, over existing alternatives, particularly in terms of the temporal consistency of the estimated depth. Additionally, we highlight the benefits of more consistent video depth in two practical applications: depth-conditioned video generation and novel view synthesis. Our project page is available at https://jhaoshao.github.io/ChronoDepth/.

We introduce a novel approach for adapting deep stereo networks in a collaborative manner. By building over principles of federated learning, we develop a distributed framework allowing for demanding the optimization process to a number of clients deployed in different environments. This makes it possible, for a deep stereo network running on resourced-constrained devices, to capitalize on the adaptation process carried out by other instances of the same architecture, and thus improve its accuracy in challenging environments even when it cannot carry out adaptation on its own. Experimental results show how federated adaptation performs equivalently to on-device adaptation, and even better when dealing with challenging environments., Comment: CVPR 2024. Project page: https://fedstereo.github.io/

Integrating an RGB camera into a ToF imaging system has become a significant technique for perceiving the real world. The RGB guided ToF imaging system is crucial to several applications, including face anti-spoofing, saliency detection, and trajectory prediction. Depending on the distance of the working range, the implementation schemes of the RGB guided ToF imaging systems are different. Specifically, ToF sensors with a uniform field of illumination, which can output dense depth but have low resolution, are typically used for close-range measurements. In contrast, LiDARs, which emit laser pulses and can only capture sparse depth, are usually employed for long-range detection. In the two cases, depth quality improvement for RGB guided ToF imaging corresponds to two sub-tasks: guided depth super-resolution and guided depth completion. In light of the recent significant boost to the field provided by deep learning, this paper comprehensively reviews the works related to RGB guided ToF imaging, including network structures, learning strategies, evaluation metrics, benchmark datasets, and objective functions. Besides, we present quantitative comparisons of state-of-the-art methods on widely used benchmark datasets. Finally, we discuss future trends and the challenges in real applications for further research., Comment: To appear on International Journal of Computer Vision (IJCV)

Generalization is a main issue for current audio deepfake detectors, which struggle to provide reliable results on out-of-distribution data. Given the speed at which more and more accurate synthesis methods are developed, it is very important to design techniques that work well also on data they were not trained for. In this paper we study the potential of large-scale pre-trained models for audio deepfake detection, with special focus on generalization ability. To this end, the detection problem is reformulated in a speaker verification framework and fake audios are exposed by the mismatch between the voice sample under test and the voice of the claimed identity. With this paradigm, no fake speech sample is necessary in training, cutting off any link with the generation method at the root, and ensuring full generalization ability. Features are extracted by general-purpose large pre-trained models, with no need for training or fine-tuning on specific fake detection or speaker verification datasets. At detection time only a limited set of voice fragments of the identity under test is required. Experiments on several datasets widespread in the community show that detectors based on pre-trained models achieve excellent performance and show strong generalization ability, rivaling supervised methods on in-distribution data and largely overcoming them on out-of-distribution data.

In this work we present an overview of approaches for the detection and attribution of synthetic images and highlight their strengths and weaknesses. We also point out and discuss hot topics in this field and outline promising directions for future research.

This paper discusses the results of the third edition of the Monocular Depth Estimation Challenge (MDEC). The challenge focuses on zero-shot generalization to the challenging SYNS-Patches dataset, featuring complex scenes in natural and indoor settings. As with the previous edition, methods can use any form of supervision, i.e. supervised or self-supervised. The challenge received a total of 19 submissions outperforming the baseline on the test set: 10 among them submitted a report describing their approach, highlighting a diffused use of foundational models such as Depth Anything at the core of their method. The challenge winners drastically improved 3D F-Score performance, from 17.51% to 23.72%., Comment: To appear in CVPRW2024

Turbulent flows over a large surface area (S) covered by n obstacles experience an overall drag due to the presence of the ground and the protruding obstacles into the flow. The drag partition between the roughness obstacles and the ground is analyzed using an analytical model proposed by Raupach (1992) and is hereafter referred to as R92. The R92 is based on the premise that the wake behind an isolated roughness element can be described by a shelter area A and a shelter volume V. The individual sizes of A and V without any interference from other obstacles can be determined from scaling analysis for the spread of wakes. To upscale from an individual roughness element to n/S elements where wakes may interact, R92 adopted a background stress re-normalizing instead of reducing A or V with each element addition. This work shows that R92's approach only converges to a linear reduction in A and V for small n/S where wakes have low probability of interacting with one another. This probabilistic nature suggests that up-scaling from individual to multiple roughness elements can be re-formulated using stochastic averaging methods proposed here. The two approaches are shown to recover R92 under plausible conditions. Comparisons between R92 and available data on blocks and vegetation-like roughness elements confirm the practical utility of R92 and its potential use in large-scale models provided the relevant parameters accommodate certain features of the roughness element type (cube versus vegetation-like) and, to a lesser extent, their configuration throughout S.

In this paper, we study extremal values for the determinant of the weighted graph Laplacian under simple nondegeneracy conditions on the weights. We derive necessary and sufficient conditions for the determinant of the Laplacian to be bounded away from zero and for the existence of a minimizing set of weights. These conditions are given both in terms of properties of random spanning trees and in terms of a type of density on graphs. These results generalize and extend the work of [7].

Deep learning methods in LiDAR-based archaeological research often leverage visualisation techniques derived from Digital Elevation Models to enhance characteristics of archaeological objects present in the images. This paper investigates the impact of visualisations on deep learning performance through a comprehensive testing framework. The study involves the use of eight semantic segmentation models to evaluate seven diverse visualisations across two study areas, encompassing five archaeological classes. Experimental results reveal that the choice of appropriate visualisations can influence performance by up to 8%. Yet, pinpointing one visualisation that outperforms the others in segmenting all archaeological classes proves challenging. The observed performance variation, reaching up to 25% across different model configurations, underscores the importance of thoughtfully selecting model configurations and LiDAR visualisations for successfully segmenting archaeological objects.

Hyperspectral data recorded from satellite platforms are often ill-suited for geo-archaeological prospection due to low spatial resolution. The established potential of hyperspectral data from airborne sensors in identifying archaeological features has, on the other side, generated increased interest in enhancing hyperspectral data to achieve higher spatial resolution. This improvement is crucial for detecting traces linked to sub-surface geo-archaeological features and can make satellite hyperspectral acquisitions more suitable for archaeological research. This research assesses the usability of pansharpened PRISMA satellite products in geo-archaeological prospections. Three pan-sharpening methods (GSA, MTF-GLP and HySure) are compared quantitatively and qualitatively and tested over the archaeological landscape of Aquileia (Italy). The results suggest that the application of pansharpening techniques makes hyperspectral satellite imagery highly suitable, under certain conditions, to the identification of sub-surface archaeological features of small and large size.

In this work, we explore the application of modulus in matroid theory, specifically, the modulus of the family of bases of matroids. This study not only recovers various concepts in matroid theory, including the strength, fractional arboricity, and principal partitions, but also offers new insights. In the process, we introduce the concept of a Beurling set. Additionally, our study revisits and provides an alternative approach to two of Edmonds's theorems related to the base packing and base covering problems. This is our stepping stone for establishing Fulkerson modulus duality for the family of bases. Finally, we provide a relationship between the base modulus of matroids and their dual matroids, and a complete understanding of the base $p$-modulus across all values of $p$.

The drag coefficient $C_d$ for a rigid and uniformly distributed rod canopy covering a sloping channel following the instantaneous collapse of a dam was examined using flume experiments. The measurements included space $x$ and time $t$ high resolution images of the water surface $h(x,t)$ for multiple channel bed slopes $S_o$ and water depths behind the dam $H_o$ along with drag estimates provided by sequential load cells. Analysis of the Saint-Venant Equation (SVE) for the front speed using the diffusive wave approximation lead to a front velocity $U_f=\sqrt{\Gamma_h 2 g \phi_v'/(C_d m D)}$, where $\Gamma_h=-\partial h/\partial x$, $g$ is the gravitational acceleration, $\phi_v'=1-\phi_v$ is fluid volume fraction per ground area, $\phi_v=m \pi D^2/4$ is the solid volume fraction per ground area, $m$ is the number of rods per ground area, and $D$ is the rod diameter. An inferred $C_d=0.4$ from the $h(x,t)$ data near the advancing front region, also confirmed by load cell measurements, is much reduced relative to its independently measured steady-uniform flow case. This finding suggests that drag reduction mechanisms associated with transients and flow disturbances are more likely to play a dominant role when compared to conventional sheltering or blocking effects on $C_d$ examined in uniform flow. The increased air volume entrained into the advancing wave front region as determined from an inflow-outflow volume balance partly explains the $C_d$ reduction from unity.

Over the past two decades, research in the field of Simultaneous Localization and Mapping (SLAM) has undergone a significant evolution, highlighting its critical role in enabling autonomous exploration of unknown environments. This evolution ranges from hand-crafted methods, through the era of deep learning, to more recent developments focused on Neural Radiance Fields (NeRFs) and 3D Gaussian Splatting (3DGS) representations. Recognizing the growing body of research and the absence of a comprehensive survey on the topic, this paper aims to provide the first comprehensive overview of SLAM progress through the lens of the latest advancements in radiance fields. It sheds light on the background, evolutionary path, inherent strengths and limitations, and serves as a fundamental reference to highlight the dynamic progress and specific challenges.

Methods for 3D reconstruction from posed frames require prior knowledge about the scene metric range, usually to recover matching cues along the epipolar lines and narrow the search range. However, such prior might not be directly available or estimated inaccurately in real scenarios -- e.g., outdoor 3D reconstruction from video sequences -- therefore heavily hampering performance. In this paper, we focus on multi-view depth estimation without requiring prior knowledge about the metric range of the scene by proposing RAMDepth, an efficient and purely 2D framework that reverses the depth estimation and matching steps order. Moreover, we demonstrate the capability of our framework to provide rich insights about the quality of the views used for prediction. Additional material can be found on our project page https://andreaconti.github.io/projects/range_agnostic_multi_view_depth., Comment: 3DV 2024 Project Page https://andreaconti.github.io/projects/range_agnostic_multi_view_depth GitHub Page https://github.com/andreaconti/ramdepth.git

This paper proposes a new framework for depth completion robust against domain-shifting issues. It exploits the generalization capability of modern stereo networks to face depth completion, by processing fictitious stereo pairs obtained through a virtual pattern projection paradigm. Any stereo network or traditional stereo matcher can be seamlessly plugged into our framework, allowing for the deployment of a virtual stereo setup that is future-proof against advancement in the stereo field. Exhaustive experiments on cross-domain generalization support our claims. Hence, we argue that our framework can help depth completion to reach new deployment scenarios., Comment: 3DV 2024. Code: https://github.com/bartn8/vppdc - Project page: https://vppdc.github.io/

The aim of this work is to explore the potential of pre-trained vision-language models (VLMs) for universal detection of AI-generated images. We develop a lightweight detection strategy based on CLIP features and study its performance in a wide variety of challenging scenarios. We find that, contrary to previous beliefs, it is neither necessary nor convenient to use a large domain-specific dataset for training. On the contrary, by using only a handful of example images from a single generative model, a CLIP-based detector exhibits surprising generalization ability and high robustness across different architectures, including recent commercial tools such as Dalle-3, Midjourney v5, and Firefly. We match the state-of-the-art (SoTA) on in-distribution data and significantly improve upon it in terms of generalization to out-of-distribution data (+6% AUC) and robustness to impaired/laundered data (+13%). Our project is available at https://grip-unina.github.io/ClipBased-SyntheticImageDetection/

This paper tackles the challenges of self-supervised monocular depth estimation in indoor scenes caused by large rotation between frames and low texture. We ease the learning process by obtaining coarse camera poses from monocular sequences through multi-view geometry to deal with the former. However, we found that limited by the scale ambiguity across different scenes in the training dataset, a na\"ive introduction of geometric coarse poses cannot play a positive role in performance improvement, which is counter-intuitive. To address this problem, we propose to refine those poses during training through rotation and translation/scale optimization. To soften the effect of the low texture, we combine the global reasoning of vision transformers with an overfitting-aware, iterative self-distillation mechanism, providing more accurate depth guidance coming from the network itself. Experiments on NYUv2, ScanNet, 7scenes, and KITTI datasets support the effectiveness of each component in our framework, which sets a new state-of-the-art for indoor self-supervised monocular depth estimation, as well as outstanding generalization ability. Code and models are available at https://github.com/zxcqlf/GasMono, Comment: ICCV 2023. Code: https://github.com/zxcqlf/GasMono

We study the robustness of the evolution of a quantum system against small uncontrolled variations in parameters in the Hamiltonian. We show that the fidelity susceptibility, which quantifies the perturbative error to leading order, can be expressed in superoperator form and use this to derive control pulses which are robust to any class of systematic unknown errors. The proposed optimal control protocol is equivalent to searching for a sequence of unitaries that mimics the first-order moments of the Haar distribution, i.e. it constitutes a 1-design. We highlight the power of our results for error resistant single- and two-qubit gates., Comment: 18 pages, 9 figures

This paper proposes a novel framework integrating the principles of active stereo in standard passive camera systems without a physical pattern projector. We virtually project a pattern over the left and right images according to the sparse measurements obtained from a depth sensor. Any such devices can be seamlessly plugged into our framework, allowing for the deployment of a virtual active stereo setup in any possible environment, overcoming the limitation of pattern projectors, such as limited working range or environmental conditions. Experiments on indoor/outdoor datasets, featuring both long and close-range, support the seamless effectiveness of our approach, boosting the accuracy of both stereo algorithms and deep networks., Comment: ICCV 2023. Code: https://github.com/bartn8/vppstereo - Project page: https://vppstereo.github.io

The ability to detect manipulated visual content is becoming increasingly important in many application fields, given the rapid advances in image synthesis methods. Of particular concern is the possibility of modifying the content of medical images, altering the resulting diagnoses. Despite its relevance, this issue has received limited attention from the research community. One reason is the lack of large and curated datasets to use for development and benchmarking purposes. Here, we investigate this issue and propose M3Dsynth, a large dataset of manipulated Computed Tomography (CT) lung images. We create manipulated images by injecting or removing lung cancer nodules in real CT scans, using three different methods based on Generative Adversarial Networks (GAN) or Diffusion Models (DM), for a total of 8,577 manipulated samples. Experiments show that these images easily fool automated diagnostic tools. We also tested several state-of-the-art forensic detectors and demonstrated that, once trained on the proposed dataset, they are able to accurately detect and localize manipulated synthetic content, even when training and test sets are not aligned, showing good generalization ability. Dataset and code are publicly available at https://grip-unina.github.io/M3Dsynth/.

Neural implicit representations have recently demonstrated compelling results on dense Simultaneous Localization And Mapping (SLAM) but suffer from the accumulation of errors in camera tracking and distortion in the reconstruction. Purposely, we present GO-SLAM, a deep-learning-based dense visual SLAM framework globally optimizing poses and 3D reconstruction in real-time. Robust pose estimation is at its core, supported by efficient loop closing and online full bundle adjustment, which optimize per frame by utilizing the learned global geometry of the complete history of input frames. Simultaneously, we update the implicit and continuous surface representation on-the-fly to ensure global consistency of 3D reconstruction. Results on various synthetic and real-world datasets demonstrate that GO-SLAM outperforms state-of-the-art approaches at tracking robustness and reconstruction accuracy. Furthermore, GO-SLAM is versatile and can run with monocular, stereo, and RGB-D input., Comment: ICCV 2023. Code: https://github.com/youmi-zym/GO-SLAM - Project Page: https://youmi-zym.github.io/projects/GO-SLAM/

The compact radio frequency negative ion source NIO1 (Negative Ion Optimization phase 1) has been designed, built and operated by Consorzio RFX and INFN-LNL in order to study and optimize the production and acceleration of H- ions in continuous operation. In 2020 Cs was evaporated in the source to increase the total extracted ion current. After an initial reduction of extracted electron to ion ratio and subsequently an increase of extracted negative ion current, the source performances progressively worsened, because of the excessive amount of Cs evaporated in the source; the extracted electron to ion ratio increased from below 1 to more than 10, while ion current density reduced from max. 67 A/m2 ion current to not more than 30 A/m2). The paper presents the experimental observations collected during Cs evaporation (reduction of plasma light, Cs emission and H$\beta$/H$\gamma$ ratio, etc.) that can help stopping the process before an excessive amount of Cs is introduced in the source. The paper also reports the cleaning techniques tested to remove the Cs excess by the action of hydrogen or argon plasmas; while argon was predictably more effective in surface sputtering, a 3 h Ar plasma treatment was not sufficient to recover from overcesiation., Comment: 6 pages, 5 figures. Accepted manuscript of a published paper

An investigation of the quasiprojectile breakup channel in semiperipheral and peripheral collisions of $^{58,64}$Ni+$^{58,64}$Ni at 32 and 52 MeV/nucleon is presented. Data have been acquired in the first experimental campaign of the INDRA-FAZIA apparatus in GANIL. The effect of isospin diffusion between projectile and target in the two asymmetric reactions has been highlighted by means of the isospin transport ratio technique, exploiting the neutron-to-proton ratio of the quasiprojectile reconstructed from the two breakup fragments. We found evidence that, for the same reaction centrality, a higher degree of relaxation of the initial isospin imbalance is achieved in the breakup channel with respect to the more populated binary output, possibly indicating the indirect selection of specific dynamical features. We have proposed an interpretation based on different average projectile-target contact times related to the two exit channels under investigation, with a longer interaction for the breakup channel. The time information has been extracted from AMD simulations of the studied systems coupled to GEMINI++: the model calculations support the hypothesis hereby presented.

In this paper, we tackle the problem of generating a novel image from an arbitrary viewpoint given a single frame as input. While existing methods operating in this setup aim at predicting the target view depth map to guide the synthesis, without explicit supervision over such a task, we jointly optimize our framework for both novel view synthesis and depth estimation to unleash the synergy between the two at its best. Specifically, a shared depth decoder is trained in a self-supervised manner to predict depth maps that are consistent across the source and target views. Our results demonstrate the effectiveness of our approach in addressing the challenges of both tasks allowing for higher-quality generated images, as well as more accurate depth for the target viewpoint.

The goal of Online Domain Adaptation for semantic segmentation is to handle unforeseeable domain changes that occur during deployment, like sudden weather events. However, the high computational costs associated with brute-force adaptation make this paradigm unfeasible for real-world applications. In this paper we propose HAMLET, a Hardware-Aware Modular Least Expensive Training framework for real-time domain adaptation. Our approach includes a hardware-aware back-propagation orchestration agent (HAMT) and a dedicated domain-shift detector that enables active control over when and how the model is adapted (LT). Thanks to these advancements, our approach is capable of performing semantic segmentation while simultaneously adapting at more than 29FPS on a single consumer-grade GPU. Our framework's encouraging accuracy and speed trade-off is demonstrated on OnDA and SHIFT benchmarks through experimental results., Comment: ICCV 2023. The first two authors contributed equally. Project page: https://marcbotet.github.io/hamlet-web/

Inferring the depth of transparent or mirror (ToM) surfaces represents a hard challenge for either sensors, algorithms, or deep networks. We propose a simple pipeline for learning to estimate depth properly for such surfaces with neural networks, without requiring any ground-truth annotation. We unveil how to obtain reliable pseudo labels by in-painting ToM objects in images and processing them with a monocular depth estimation model. These labels can be used to fine-tune existing monocular or stereo networks, to let them learn how to deal with ToM surfaces. Experimental results on the Booster dataset show the dramatic improvements enabled by our remarkably simple proposal., Comment: Accepted at ICCV 2023. Project Page: https://cvlab-unibo.github.io/Depth4ToM

This work presents a new unsupervised framework for training deep learning models for super-resolution of Sentinel-2 images by fusion of its 10-m and 20-m bands. The proposed scheme avoids the resolution downgrade process needed to generate training data in the supervised case. On the other hand, a proper loss that accounts for cycle-consistency between the network prediction and the input components to be fused is proposed. Despite its unsupervised nature, in our preliminary experiments the proposed scheme has shown promising results in comparison to the supervised approach. Besides, by construction of the proposed loss, the resulting trained network can be ascribed to the class of multi-resolution analysis methods.