Thousands of readers who have profited from engagement with the lively mind of Rudolf Arnheim over the decades will receive news of this new collection of essays expectantly. In the essays collected here, as in his earlier work on a large variety of art forms, Arnheim explores concrete poetry and the metaphors of Dante, photography and the meaning of music. There are essays on color composition, forgeries, and the problems of perspective, on art in education and therapy, on the style of artists' late works, and the reading of maps. Also, in a triplet of essays on pioneers in the psychology of art (Max Wertheimer, Gustav Theodor Fechner, and Wilhelm Worringer) Arnheim goes back to the roots of modern thinking about the mechanisms of artistic perception.
Embodied agents for creative tasks like photography must bridge the semantic gap between high-level language commands and geometric control. We introduce PhotoAgent, an agent that achieves this by integrating Large Multimodal Models (LMMs) reasoning with a novel control paradigm. PhotoAgent first translates subjective aesthetic goals into solvable geometric constraints via LMM-driven, chain-of-thought (CoT) reasoning, allowing an analytical solver to compute a high-quality initial viewpoint. This initial pose is then iteratively refined through visual reflection within a photorealistic internal world model built with 3D Gaussian Splatting (3DGS). This ``mental simulation'' replaces costly and slow physical trial-and-error, enabling rapid convergence to aesthetically superior results. Evaluations confirm that PhotoAgent excels in spatial reasoning and achieves superior final image quality.
The Sun has been observed through a telescope for four centuries. However, its study made a prodigious leap at the end of the nineteenth century with the appearance of photography and spectroscopy, then at the beginning of the following century with the invention of the coronagraph and monochromatic filters, and finally in the second half of the twentieth century with the advent of large ground-based telescopes and space exploration. This article retraces the main stages of solar instrumental developments in Meudon, from its foundation by Jules Janssen in 1876 to the present day, limited to ground-based or balloon instrumentation, designed in Meudon and installed there or in other places (Nan{\c c}ay, Pic du Midi, Canary Islands). The Meudon astronomers played a pioneering role in the history of solar physics through the experimentation of innovative techniques. After the golden age of inventions, came the time of large instruments, studied in Meudon but often installed in more favourable sites, and that of space, in a framework of international collaboration, but this is not discussed here.
Paola Mattioli's work in fashion photography is characterized by keywords such as "portrait," "canon," and "representation of women." Through collaborations with prominent magazines like "Amica" and the feminist publication "Grattacielo: Women's Eyes on the World," Mattioli establishes herself as a key figure in feminist photographic practices. Transitioning from socio-cultural reflections of the late '60s, she began publishing fashion spreads in Italian magazines in the '80s, continuing into the 2000s, culminating in collaborations with Christian Dior. Mattioli's fashion photography during the 1981-1991 period encompasses two types of projects: collaborations with magazines and direct partnerships with brands like Enrica Massei and Nanni Strada. She delineates two approaches to her fashion photography: one focused on portraiture and the other on scene creation, both aimed at redefining cultural models, particularly those of femininity. A critical exploration of fashion photography requires moving beyond viewing fashion as mere clothing, as it embodies attitudes, linguistic forms, and lifestyle choices. Fashion serves as an identity vehicle, contributing to identity formation, as discussed by Lars Svendsen and Roland Barthes. Mattioli engages in sociological and aesthetic perspectives, particularly addressing the concept of identity in relation to subjectivity. Operating in a predominantly male-dominated context in fashion photography, Mattioli challenges the male gaze and redefines female representation through her lens. Her work intersects with feminist discourse, in line with Eugenia Paulicelli's view that fashion provides a vital space for identity construction and resistance against stereotypical images. She consciously subverts traditional fashion poses, offering new perspectives rooted in feminist concepts.
Apple cultivar classification is challenging due to the inter-class similarity and high intra-class variations. Human experts do not rely on single-view features but rather study each viewpoint of the apple to identify a cultivar, paying close attention to various details. Following our previous work, we try to establish a similar multiview approach for machine-learning (ML)-based apple classification in this paper. In our previous work, we studied apple classification using one single view. While these results were promising, it also became clear that one view alone might not contain enough information in the case of many classes or cultivars. Therefore, exploring multiview classification for this task is the next logical step. Multiview classification is nothing new, and we use state-of-the-art approaches as a base. Our goal is to find the best approach for the specific apple classification task and study what is achievable with the given methods towards our future goal of applying this on a mobile device without the need for internet connectivity. In this study, we compare an ensemble model with two cases where we use single networks: one without view specialization trained on all available images without view assignment and one where we combine the separate views into a single image of one specific instance. The two latter options reflect dataset organization and preprocessing to allow the use of smaller models in terms of stored weights and number of operations than an ensemble model. We compare the different approaches based on our custom apple cultivar dataset. The results show that the state-of-the-art ensemble provides the best result. However, using images with combined views shows a decrease in accuracy by 3% while requiring only 60% of the memory for weights. Thus, simpler approaches with enhanced preprocessing can open a trade-off for classification tasks on mobile devices.
Photography, Computer applications to medicine. Medical informatics
The Sun has been observed through a telescope for four centuries. However, its study made a prodigious leap at the end of the nineteenth century with the appearance of photography and spectroscopy, then at the beginning of the following century with the invention of the coronagraph and monochromatic filters, and finally in the second half of the twentieth century with the advent of space exploration (satellites, probes). This makes it possible to observe the radiations hidden by the Earth's atmosphere (Ultra Violet, X-rays, $γ$) and to carry out ''in situ'' measurements in the solar environment. This article retraces the major stages of this fantastic epic in which renowned scientists such as Janssen, Deslandres, d'Azambuja, Lyot and Dollfus entered the scene, giving the Paris-Meudon Observatory a pioneering role in the history of solar physics until 1960. After this golden age, space exploration required large resources shared between nations, which could no longer be implemented within teams or even individual institutes. The development of numerical simulation, a new research tool, also required the pooling of supercomputers.
The low dynamic range (LDR) of common cameras fails to capture the rich contrast in natural scenes, resulting in loss of color and details in saturated pixels. Reconstructing the high dynamic range (HDR) of luminance present in the scene from single LDR photographs is an important task with many applications in computational photography and realistic display of images. The HDR reconstruction task aims to infer the lost details using the context present in the scene, requiring neural networks to understand high-level geometric and illumination cues. This makes it challenging for data-driven algorithms to generate accurate and high-resolution results. In this work, we introduce a physically-inspired remodeling of the HDR reconstruction problem in the intrinsic domain. The intrinsic model allows us to train separate networks to extend the dynamic range in the shading domain and to recover lost color details in the albedo domain. We show that dividing the problem into two simpler sub-tasks improves performance in a wide variety of photographs.
Unmanned aerial vehicle (UAV) tilt photography technology has gradually become a new technical means of disaster risk identification. This technology combines UAVs, satellite remote sensing, and ground online monitoring systems to establish an integrated space–sky–Earth system that can be used for tailings pond risk identification. With the use of this system for visual interpretation, water body identification, and monitoring data analysis, multiple types of monitoring parameters of a typical tailings pond in China, such as the seepage line and surface deformation, were obtained. Moreover, intelligent fusion analysis was performed of multisource data to outline the problems affecting tailings safety in the process of elevation expansion and irregular ore discharge of the tailings pond. Warning values of different levels were obtained to assess the overall safety condition of the tailings pond, and the proposed technology was verified. The research results could provide a new basis for accurate evaluation of the running state of tailings ponds and offer an effective remote monitoring means for tailings pond enterprises and supervisory departments.
Despite the significant progress in 6-DoF visual localization, researchers are mostly driven by ground-level benchmarks. Compared with aerial oblique photography, ground-level map collection lacks scalability and complete coverage. In this work, we propose to go beyond the traditional ground-level setting and exploit the cross-view localization from aerial to ground. We solve this problem by formulating camera pose estimation as an iterative render-and-compare pipeline and enhancing the robustness through augmenting seeds from noisy initial priors. As no public dataset exists for the studied problem, we collect a new dataset that provides a variety of cross-view images from smartphones and drones and develop a semi-automatic system to acquire ground-truth poses for query images. We benchmark our method as well as several state-of-the-art baselines and demonstrate that our method outperforms other approaches by a large margin.
Abstract Background Triple-negative breast cancer (TNBC) is an aggressive subtype of breast cancer and is susceptible to develop gemcitabine (GEM) resistance. Decreased expression of human equilibrative nucleoside transporter 1 (hENT1) accompanied by compensatory increase of glycolysis is strongly associated with GEM resistance in TNBC. In this study, we investigated the treatment feasibility of combined hENT1 upregulation and miR-143-mediated inhibition of glycolysis for reversing GEM resistance in TNBC. Methods Experiments were performed in vitro and in vivo to compare the efficacy of GEM therapies. In this study, we established stable drug-resistant cell line, GEM-R cells, from parental cells (MDA-MB-231) through exposure to GEM following a stepwise incremental dosing strategy. Then GEM-R cells were transfected by lentiviral plasmids and GEM-R cells overexpressing hENT1 (GEM-R-hENT1) were established. The viability and apoptosis of wild-type (MDA-MB-231), GEM-R, and GEM-R-hENT1 cells treated with GEM or GEM + miR-143 were analyzed by CCK8 assay and flow cytometry. The RNA expression and protein expression were measured by RT-PCR and western blotting respectively. GEM uptake was determined by multiple reaction monitoring (MRM) analysis. Glycolysis was measured by glucose assay and 18F-FDG uptake. The antitumor effect was assessed in vivo in a tumor xenograft model by evaluating toxicity, tumor volume, and maximum standardized uptake value in 18F-FDG PET. Immunohistochemistry and fluorescence photography were taken in tumor samples. Pairwise comparisons were performed using Student’s t-test. Results Our results represented that overexpression of hENT1 reversed GEM resistance in GEM-R cells by showing lower IC50 and higher rate of apoptosis. MiR-143 suppressed glycolysis in GEM-R cells and enhanced the effect of reversing GEM resistance in GEM-R-hENT1 cells. The therapeutic efficacy was validated using a xenograft mouse model. Combination treatment decreased tumor growth rate and maximum standardized uptake value in 18F-FDG PET more effectively. Conclusions Combined therapy of exogenous upregulation of hENT1 expression and miR-143 mimic administration was effective in reversing GEM resistance, providing a promising strategy for treating GEM-resistant TNBC.
Neoplasms. Tumors. Oncology. Including cancer and carcinogens, Cytology
George Braid, Carlota Ruiz de Galarreta, Andrew Comley
et al.
The control of a lens’s numerical aperture has potential applications in areas such as photography and imaging, displays, sensing, laser processing and even laser-implosion fusion. In such fields, the ability to control lens properties dynamically is of much interest, and active meta-lenses of various kinds are under investigation due to their modulation speed and compactness. However, as of yet, meta-lenses that explicitly offer dynamic control of a lens’s numerical aperture have received little attention. Here, we design and simulate active meta-lenses (specifically, focusing meta-mirrors) using chalcogenide phase-change materials to provide such control. We show that, operating at a wavelength of 3000 nm, our devices can change the numerical aperture by up to a factor of 1.85 and operate at optical intensities of the order of 1.2 × 10<sup>9</sup> Wm<sup>−2</sup>. Furthermore, we show the scalability of our design towards shorter wavelengths (visible spectrum), where we demonstrate a change in NA by a factor of 1.92.
Abstract With the fast development of deep learning models, hierarchical convolutional neural networks have achieved great success in image denoising tasks. To further boost the performance of image denoising, a novel non‐local hierarchical network (NHNet) is proposed. Unlike existing U‐Net‐based hierarchical methods, which mainly focus on downsampling operations, NHNet adopts an initial resolution path and a high resolution path. Specifically, the high‐resolution features are obtained through upsampling, where the non‐local mechanism is adopted to capture the self‐similarity properties, which contribute to a better denoising performance. Cross connections and channel attention layers are added between the two paths to integrate features in different resolutions. Compared with other U‐Net‐based hierarchical networks, NHNet requires fewer parameters. Experiments show that NHNet achieves state‐of‐the‐art performance in Gaussian denoising tasks and gets competitive results when dealing with real image denoising.
HDR is an important part of computational photography technology. In this paper, we propose a lightweight neural network called Efficient Attention-and-alignment-guided Progressive Network (EAPNet) for the challenge NTIRE 2022 HDR Track 1 and Track 2. We introduce a multi-dimensional lightweight encoding module to extract features. Besides, we propose Progressive Dilated U-shape Block (PDUB) that can be a progressive plug-and-play module for dynamically tuning MAccs and PSNR. Finally, we use fast and low-power feature-align module to deal with misalignment problem in place of the time-consuming Deformable Convolutional Network (DCN). The experiments show that our method achieves about 20 times compression on MAccs with better mu-PSNR and PSNR compared to the state-of-the-art method. We got the second place of both two tracks during the testing phase. Figure1. shows the visualized result of NTIRE 2022 HDR challenge.
We introduce 3D Moments, a new computational photography effect. As input we take a pair of near-duplicate photos, i.e., photos of moving subjects from similar viewpoints, common in people's photo collections. As output, we produce a video that smoothly interpolates the scene motion from the first photo to the second, while also producing camera motion with parallax that gives a heightened sense of 3D. To achieve this effect, we represent the scene as a pair of feature-based layered depth images augmented with scene flow. This representation enables motion interpolation along with independent control of the camera viewpoint. Our system produces photorealistic space-time videos with motion parallax and scene dynamics, while plausibly recovering regions occluded in the original views. We conduct extensive experiments demonstrating superior performance over baselines on public datasets and in-the-wild photos. Project page: https://3d-moments.github.io/
Striking pictures showing filamentary structures in plasma focus have intrigued researchers from the early days of plasma focus research. A definitive understanding of their occurrence, origin, structure and role in plasma focus physics is still not in sight as summarized in a recent comprehensive review. This is because they are often not observed in a "standard mode" of plasma focus operation with pure deuterium, particularly in large installations, but are found in smaller experiments or those with gaseous admixtures. This has led to the suspicion that filaments are not a native feature of the plasma focus phenomenon. Recent success in observation of filaments in PF-1000 in a pure deuterium operation by a novel modification of the interferometer system that allows simultaneous interferometry and schlieren photography changes this situation. This Letter looks at the implications of this development in the larger context of plasma focus physics. Conceptualization of filamentation as a native feature of the traveling current distribution behind an ionizing strong shock wave is shown to be a feasible paradigm that can be formulated as a computable model for filamentation in the plasma focus.
As humans, we can remember certain visuals in great detail, and sometimes even after viewing them once. What is even more interesting is that humans tend to remember and forget the same things, suggesting that there might be some general internal characteristics of an image to encode and discard similar types of information. Research suggests that some pictures tend to be memorized more than others. The ability of an image to be remembered by different viewers is one of its intrinsic properties. In visualization and photography, creating memorable images is a difficult task. Hence, to solve the problem, various techniques predict visual memorability and manipulate images' memorability. We present a comprehensive literature survey to assess the deep learning techniques used to predict and modify memorability. In particular, we analyze the use of Convolutional Neural Networks, Recurrent Neural Networks, and Generative Adversarial Networks for image memorability prediction and modification.
A shallow depth-of-field image keeps the subject in focus, and the foreground and background contexts blurred. This effect requires much larger lens apertures than those of smartphone cameras. Conventional methods acquire RGB-D images and blur image regions based on their depth. However, this approach is not suitable for reflective or transparent surfaces, or finely detailed object silhouettes, where the depth value is inaccurate or ambiguous. We present a learning-based method to synthesize the defocus blur in shallow depth-of-field images from handheld bursts acquired with a single small aperture lens. Our deep learning model directly produces the shallow depth-of-field image, avoiding explicit depth-based blurring. The simulated aperture diameter equals the camera translation during burst acquisition. Our method does not suffer from artifacts due to inaccurate or ambiguous depth estimation, and it is well-suited to portrait photography.
With the advent of deep learning methods replacing the ISP in transforming sensor RAW readings into RGB images, numerous methodologies solidified into real-life applications. Equally potent is the task of inverting this process which will have applications in enhancing computational photography tasks that are conducted in the RAW domain, addressing lack of available RAW data while reaping from the benefits of performing tasks directly on sensor readings. This paper's proposed methodology is a state-of-the-art solution to the task of RAW reconstruction, and the multi-step refinement process integrating an overexposure mask is novel in three ways: instead of from RGB to bayer, the pipeline trains from RGB to demosaiced RAW allowing use of perceptual loss functions; the multi-step processes has greatly enhanced the performance of the baseline U-Net from start to end; the pipeline is a generalizable process of refinement that can enhance other high performance methodologies that support end-to-end learning.