Specific Scene Research Articles

Stealthy Vehicle Adversarial Camouflage Texture Generation Based on Neural Style Transfer

Adversarial attacks that mislead deep neural networks (DNNs) into making incorrect predictions can also be implemented in the physical world. However, most of the existing adversarial camouflage textures that attack object detection models only consider the effectiveness of the attack, ignoring the stealthiness of adversarial attacks, resulting in the generated adversarial camouflage textures appearing abrupt to human observers. To address this issue, we propose a style transfer module added to an adversarial texture generation framework. By calculating the style loss between the texture and the specified style image, the adversarial texture generated by the model is guided to have good stealthiness and is not easily detected by DNNs and human observers in specific scenes. Experiments have shown that in both the digital and physical worlds, the vehicle full coverage adversarial camouflage texture we create has good stealthiness and can effectively fool advanced DNN object detectors while evading human observers in specific scenes.

Standing Still, Go Along, and Zooming In: Complementary Strategies of Participant-Observations in Urban Ethnography

This paper seeks to complement conventional ethnographic participant observation methods, proposing the integration of three distinct strategies. Motivated by own ethnographic studies, conducted in an urban supermarket and a food festival site, the paper contends that traditional long-term immersion may not always suffice for investigating transient urban phenomena. The paper advocates for the integration of complementary methodological strategies for participant observation adaptable to evolving contexts and research questions. Firstly, the "standing still" approach involves minimal engagement while observing a specific scene or interaction, offering a detached yet insightful perspective on fleeting moments or events. Secondly, the "go along" strategy entails tracking a particular individual, group, or object through movements and networks, providing valuable insights into routines, relationships, and controversies. Lastly, the "zooming in" technique uncovers unexpected details within scenes, interactions, or subject matter, enriching understanding of what might otherwise go unnoticed. While the first two strategies are practices employed by early pioneers in the field, the latter gains relevance in an era where video recordings are increasingly prevalent. The paper argues for the complementarity of the three strategies, fostering a more advanced comprehension of the supermarket and social landscape at large. This study holds significant implications for the practice of participant observation, encouraging researchers to embrace a repertoire of strategies tailored to the unique demands of each research inquiry and situational context. This attentive and context-driven approach allows for a more nuanced and comprehensive grasp of the social world under scrutiny.

Choice cuts for Khakaura: New Kingdom Captions in the North Chapel of Senwosret III, Dahshur

The paper analyses a group of New Kingdom hieratic ink-written labels added to the Middle Kingdom relief figures of offering bearers in the North Chapel of the pyramid precinct of King Senwosret III of the 12th Dynasty at Dahshur. It applies a close observation of the spatial properties of the secondary epigraphy combined with contexts of the Egyptian written culture. To a large extent, it is a beginning, rather than a fixed and detailed outline, of a more differentiated, location-specific interpretation of secondary texts and figures added to the decoration of various parts of the royal pyramid complexes. The interpretation of this specific scene suggests that the visitors articulated their status alongside a personal involvement with this prestigious site by labelling the offering bearers in order to achieve a purposeful revivification, and an arrogation, of the offering scene performance. [Formula: see text]

Scene understanding method utilizing global visual and spatial interaction features for safety production

Risk identification in power operations is crucial for both personal safety and power production. Existing risk identification methods mainly use target detection models to identify the common risks but the scene specificity of risk occurrence. For example, not wearing a safety harness, not wearing insulated gloves, etc. Since most methods for detecting safety gears make sense only under specific scene. But the power electric work is a complex object involving many elements such as personnel, equipment and safety tools. Therefore, this paper proposes a scene understanding method that integrates visual features and spatial relationship features among scene elements. This method constructs a scenean undirected scene graph to represent the interactive relationship among the elements, extracts the interactive features by using a graph encoder-decoder convolution module, and fuse perceived high-dimensional visual features and spatial topological features for scene recognition, in order to effectively solve addressing the power operation scene understanding problem under multi-element interaction. Finally, a power inspection operation scenario was chosen as the test case. The outcome of the evaluation indicates results indicate that the proposed approach suggested in this study exhibits superior precision in scene identification and shows ademonstrates strong generalization ability.

The Types of Discrimination and Princess Diana’s Ways to Pursue Her Freedom in Spencer (2021)

This article elaborates the film "Spencer," characterized as "a fable from true tragedy" based on Princess Diana's story. The focus of the research is on how Princess Diana pursued her freedom. The methodology involves a direct viewing and interpretation of "Spencer," analyzing visual scenes and audio dialogues. Following Roland Barthes' semiotic approach, the researcher identifies denotative signs, transitions them into connotative markers, and ultimately explores the mythic stage. Specific scenes, such as Princess Diana encountering a sign urging silence and her secretive actions in the palace, are dissected to reveal layers of meaning. Through semiotic analysis, the article illustrates Princess Diana's attempts to break free from royal constraints. Each scene reflects the conflict between individual desires for freedom and the norms imposed by the royal institution. The study offers a nuanced understanding of Princess Diana's struggle within the royal family's rules and pressures. The analysis provides profound insights into how "Spencer" portrays Princess Diana's quest for personal freedom amidst the rigid regulations of the royal household, employing semiotics as a valuable approach to unravel the meanings embedded in visual and auditory elements of the film.

The visual statistical learning overcomes scene dissimilarity through an independent clustering process.

Contextual cueing is a phenomenon of visual statistical learning observed in visual search tasks. Previous research has found that the degree of deviation of items from its centroid, known as variability, determines the extent of generalization for that repeated scene. Introducing variability increases dissimilarity between multiple occurrences of the same repeated layout significantly. However, current theories do not explain the mechanisms that help to overcome this dissimilarity during contextual cue learning. We propose that the cognitive system initially abstracts specific scenes into scene layouts through an automatic clustering unrelated to specific repeated scenes, and subsequently uses these abstracted scene layouts for contextual cue learning. Experiment 1 indicates that introducing greater variability in search scenes leads to a hindering in the contextual cue learning. Experiment 2 further establishes that conducting extensive visual searches involving spatial variability in entirely novel scenes facilitates subsequent contextual cue learning involving corresponding scene variability, confirming that learning clustering knowledge precedes the contextual cue learning and is independent of specific repeated scenes. Overall, this study demonstrates the existence of multiple levels of learning in visual statistical learning, where item-level learning can serve as material for layout-level learning, and the generalization reflects the constraining role of item-level knowledge on layout-level knowledge.

Ground-to-air aircraft infrared image deblurring based on imaging degradation simulation

The issue of infrared image deblurring has been a significant concern. However, in some specific scenes, the current mainstream deblurring algorithms based on optimization or deep learning fail to provide satisfactory results. Aiming to address the ineffectiveness of deep learning methods due to the low-cost datasets' unavailability for specific scenes, we innovatively propose a relatively simple full-chain imaging degradation simulation method using ground-to-air aircraft infrared imaging scene as an example, which considers the effects of blur and noise caused by the atmosphere, imaging system, target motion and detector. Through this method, we could generate abundant blur-clear image pairs by altering various parameters. To enhance the neural network’s generalization ability and the deblurring performance in the specific scenes, we employ a two-step approach: pretraining on the public GoPro dataset and subsequent finetuning on the simulation dataset. After testing on the simulation dataset and some real-world images, we have discovered the importance of selecting a pretraining dataset that closely matches the scene degradation mode. Additionally, regardless of whether the model is pre-trained on the UIRD or GoPro dataset, there are significant enhancements in the deblurring effect following finetuning with our constructed simulation dataset. In summary, compared to the traditional deconvolution methods and the methods trained on a general dataset, our approach not only exhibits superior deblurring capabilities but also effectively mitigates noise and prevents the occurrence of artifactual textures such as ringing artifact.

Efficient greenhouse segmentation with visual foundation models: achieving more with fewer samples

Introduction: The Vision Transformer (ViT) model, which leverages self-supervised learning, has shown exceptional performance in natural image segmentation, suggesting its extensive potential in visual tasks. However, its effectiveness diminishes in remote sensing due to the varying perspectives of remote sensing images and unique optical properties of features like the translucency of greenhouses. Additionally, the high cost of training Visual Foundation Models (VFMs) from scratch for specific scenes limits their deployment.Methods: This study investigates the feasibility of rapidly deploying VFMs on new tasks by using embedding vectors generated by VFMs as prior knowledge to enhance traditional segmentation models’ performance. We implemented this approach to improve the accuracy and robustness of segmentation with the same number of trainable parameters. Comparative experiments were conducted to evaluate the efficiency and effectiveness of this method, especially in the context of greenhouse detection and management.Results: Our findings indicate that the use of embedding vectors facilitates rapid convergence and significantly boosts segmentation accuracy and robustness. Notably, our method achieves or exceeds the performance of traditional segmentation models using only about 40% of the annotated samples. This reduction in the reliance on manual annotation has significant implications for remote sensing applications.Discussion: The application of VFMs in remote sensing tasks, particularly for greenhouse detection and management, demonstrated enhanced segmentation accuracy and reduced dependence on annotated samples. This method adapts more swiftly to different lighting conditions, enabling more precise monitoring of agricultural resources. Our study underscores the potential of VFMs in remote sensing tasks and opens new avenues for the expansive application of these models in diverse downstream tasks.

Product Placement Dalam Serial Drama Korea (Analisis Isi Kualitatif terhadap Evaluasi Atribut Product Placement Kopiko dalam Drama Korea Vincenzo)

This study aims to analyze the implementation of Kopiko product placement in the Korean drama "Vincenzo" through Russell's (2002) dimensions of product placement, while evaluating the associated attributes of Kopiko brand placement within it. Considering the attributes of Entertainment, Informativeness, and Discomfort as outlined by Ducoffe (1996), the study adopts a qualitative approach with a constructivist paradigm, utilizing content analysis as the research method. The units of analysis include episodes 14, 15, 17, and 19 of the drama "Vincenzo". The analysis results indicate that the product placement strategy successfully optimizes visual, entertainment, and informative attributes in each episode. Aligning product messages with the storyline is crucial to maintaining narrative integrity without disrupting the viewer's experience. For instance, dialogues promoting the consumption of Kopiko candies in specific scenes demonstrate effective alignment between product messaging and narrative context. The informative attribute of product placement is also demonstrated effectively, particularly in episode 19 where information about Kopiko's product advantages is conveyed through dialogue, providing added value to viewers with relevant product information aligned with product placement attributes. Nevertheless, there is a risk of discomfort from product placement if the integration of the product within the storyline is inadequate. This study underscores the importance of considering product integration with a consistent narrative to avoid disruptions to the viewer's experience.

Oppression and Females’ Defending in the Devil Wears Prada

Although the film industry has improved in terms of portraying female characters since the 1960s, there are still issues visible to the public, and one is the hidden objectification of females in supposedly feminist movies. Another popular opinion recently, especially on Chinese social media platforms, is that fashion acts as a tool of the patriarchy to force the male gaze onto women. However, this paper intends to discuss how the fashion industry acts as a non-patriarchal field that provides opportunities for females in terms of improving themselves. The movie, The Devil Wears Prada, will be analyzed as an example to explain how the fashion industry, in this movie, does not consist of the male gaze and propels the main character’s growth. The approaches used in this paper are Laura Mulvey’s gaze theory and analyzing specific scenes in the movie, The Devil Wears Prada. The scenes will be analyzed using the apparatus theory in films, explaining how the different types of shots influence the message conveyed in the movie. In conclusion, this paper reveals that even a feminist film like The Devil Wears Prada contains various circumstances of the male gaze; however, these circumstances of the male gaze influencing women do not appear because of the fashion industry, and consequently, the fashion industry should not be disapproved as a source of the male gaze, as it provides opportunities for women to improve themselves in terms of their career and wellbeing.

Novel Deep Learning Domain Adaptation Approach for Object Detection Using Semi-Self Building Dataset and Modified YOLOv4

Moving object detection is a vital research area that plays an essential role in intelligent transportation systems (ITSs) and various applications in computer vision. Recently, researchers have utilized convolutional neural networks (CNNs) to develop new techniques in object detection and recognition. However, with the increasing number of machine learning strategies used for object detection, there has been a growing need for large datasets with accurate ground truth used for the training, usually demanding their manual labeling. Moreover, most of these deep strategies are supervised and only applicable for specific scenes with large computational resources needed. Alternatively, other object detection techniques such as classical background subtraction need low computational resources and can be used with general scenes. In this paper, we propose a new a reliable semi-automatic method that combines a modified version of the detection-based CNN You Only Look Once V4 (YOLOv4) technique and background subtraction technique to perform an unsupervised object detection for surveillance videos. In this proposed strategy, background subtraction-based low-rank decomposition is applied firstly to extract the moving objects. Then, a clustering method is adopted to refine the background subtraction (BS) result. Finally, the refined results are used to fine-tune the modified YOLO v4 before using it in the detection and classification of objects. The main contribution of this work is a new detection framework that overcomes manual labeling and creates an automatic labeler that can replace manual labeling using motion information to supply labeled training data (background and foreground) directly from the detection video. Extensive experiments using real-world object monitoring benchmarks indicate that the suggested framework obtains a considerable increase in mAP compared to state-of-the-art results on both the CDnet 2014 and UA-DETRAC datasets.

A Coordinating Small Organic Molecule with Tunable Lower Critical Solution Temperature for Efficient Management of Solar Radiation.

Structurally well-defined small molecules with lower critical solution temperature (LCST) behavior offer enormous prospects for fine-tuning their phase transition properties to be "on-demand" applied in the specific scene but are still underexplored. Herein, a novel amphiphilic small LCST molecule is rationally designed and synthesized. The molecule, namely TG, features a conjugation of multiple short ethylene glycol (EG) chains with the functional coordinating terpyridine (Tpy) moiety. The molecule TG demonstrates excellent LCST behavior down to 0.05× 10-3 m in a water solution. And a cloud point Tcp = 30.9°C with a very short thermal hysteresis ΔT = 0.2°C and good reversibility can be achieved when c = 0.1× 10-3 m. The excellent LCST properties of TG have enabled its successful performance as the smart window for solar radiation management with the ∆Tlum, ∆TIR, and ∆Tsol being 83.6%, 49.1%, and 67.2%, respectively. Moreover, the presence of Tpy moiety in TG enables its coordination with Ru3+ and the resulting complex also exhibits modulated LCST behavior with different concentration-dependent Tcp. These studies would provide novel small-molecule-based scaffolds for constructing better solar radiation management systems as well as other thermal-responsive smart materials.

Analysis of Two Kinds of MV Trailer Creation Content

MV trailer is the MV inside the more eye-catching or eye-catching, to attract everyone's attention, it is a very effective way to promote the MV. This paper focuses on two different types of MV trailer type creation features, and analyzes the visual effect of MV trailer to attract the public with theoretical and practical significance. Using two representative female groups thumb amount of the first five MV forecast data in this paper found: (1) dance type MV forecast play characteristic is the music of chorus accompaniment with a short dance and different perspective scene to catch the public eye; (2) Story type MV forecast is characterized by some specific scene and perspective transformation suggested that the subsequent MV story. This paper aims to analyze the theoretical and practical significance of the visual effects of MV trailers in attracting the public by studying the creative characteristics of different types of MV trailers, and provide reference for the subsequent MV releases for better publicity.

Unveiling Voices and Empowering Narratives: A Comprehensive Exploration of Islamic Feminism in Marjane Satrapi's Persepolis

This literature paper delves into the intricate layers of Islamic feminism within Marjane Satrapi's graphic novel Persepolis. The narrative unfolds against the backdrop of post-revolutionary Iran, providing a nuanced examination of the lives of Iranian women amid political and social transformations. Through a thorough analysis of female characters, with a focus on Marjane's experiences and those of her counterparts, the paper explores instances of resistance and agency as women navigate societal expectations. The concept of Islamic Feminism is introduced and contextualized, tracing its evolution, key principles, and objectives, while highlighting its role in challenging traditional gender norms within Islamic values. The historical and cultural background of Iran during the depicted period, encompassing the Iranian Revolution and its aftermath, sets the stage for feminist discourse.Examining specific scenes and dialogues, the paper identifies and analyses elements in the narrative that reflect Islamic Feminist ideas, illustrating how Satrapi skilfully navigates the intersection of feminism and Islamic culture. Challenges and criticisms surrounding the application of the Islamic Feminism label to Persepolis are addressed, considering perspectives from both proponents and sceptics. The study concludes by discussing the potential impact of Persepolis on readers' understanding of Islamic Feminism and its broader contribution to conversations about feminism within Islamic contexts. Through a comprehensive examination, it highlights the significance of embracing diverse perspectives and employing dynamic approaches to comprehend the complexities of women's experiences within the Islamic context. By integrating ijtihad and pluralist, liberal-rationalist, and contextual approaches, it aims to unveil voices and empower narratives, ultimately fostering gender equality and justice for all. This abstract encapsulates the essence of the findings, offering insights into the multifaceted nature of Islamic feminism as pictured in Persepolis.

A weight induced contrast map for infrared and visible image fusion

Fusion involves merging details from both infrared (IR) and visible images to generate a unified composite image that offers richer and more valuable information than either individual image. Surveillance, navigation, remote sensing, and military applications require various imaging modalities, including visible and IR, to oversee specific scenes. These sensors provide supplementary data and improve situational understanding, so it is essential to fuse this information into a single image. Fusing IR and visible images presents several challenges due to the differences in imaging modalities, data characteristics, and the need for accurate and meaningful integration of information. In this context, a novel image fusion architecture focuses on enhancing prominent targets, with the objective of integrating thermal information from infrared images into visible images while preserving textural details within the visible images. Initially, in the proposed algorithm, the images from different sensors are divided into components of high and low frequencies using a Guided filter and an Average filter, respectively. A unique contrast detection mechanism is proposed that is capable of preserving the contrast information from the original images. Further, the contrast details of the IR and visible images are enhanced using local standard deviation filtering and local range filtering, respectively. We have developed a new weight map construction strategy that can effectively preserve the supplemental data from both the original images. These weights and gradient details of the source images are utilized to preserve the salient feature details of the images acquired from the various modalities. A decision-making approach is utilized among the high-frequency components of the original images to retain the prominent feature details of the source images. Finally, the salient feature details and the prominent feature details are integrated to generate the fused image. The developed technique is validated using both subjective and quantitative perspectives. The developed approaches provide EN, MI, Nabf, and SD of 6.86815, 13.73269, 0.15390, and 78.16158 respectively against deep learning-based approaches. Also, the proposed algorithm provides EN, MI, Nabf, FMIw, and Qabf against 6.86815, 13.73269, 0.15390, 0.41634 and 0.47196 respectively against existing traditional fusion methods. It is observed that the developed technique provides adequate accuracy against twenty-seven state-of-the-art techniques.

Easily deployable real-time detection method for small traffic signs

Traffic sign detection technology plays an important role in driver assistance systems and automated driving systems. This paper proposes DeployEase-YOLO, a real-time high-precision detection scheme based on an adaptive scaling channel pruning strategy, to facilitate the deployment of detectors on edge devices. More specifically, based on the characteristics of small traffic signs and complex background, this paper first of all adds a small target detection layer to the basic architecture of YOLOv5 in order to improve the detection accuracy of small traffic signs.Then, when capturing specific scenes with large fields of view, higher resolution and richer pixel information are preserved instead of directly scaling the image size. Finally, the network structure is pruned and compressed using an adaptive scaling channel pruning strategy, and the pruned network is subjected to a secondary sparse pruning operation. The number of parameters and computations is greatly reduced without increasing the depth of the network structure or the influence of the input image size, thus compressing the model to the minimum within the compressible range. Experimental results show that the model trained by Experimental results show that the model trained by DeployEase-YOLO achieves higher accuracy and a smaller size on TT100k, a challenging traffic sign detection dataset. Compared to existing methods, DeployEase-YOLO achieves an average accuracy of 93.3%, representing a 1.3% improvement over the state-of-the-art YOLOv7 network, while reducing the number of parameters and computations to 41.69% and 59.98% of the original, respectively, with a compressed volume of 53.22% of the previous one. This proves that the DeployEase-YOLO has a great deal of potential for use in the area of small traffic sign detection. The algorithm outperforms existing methods in terms of accuracy and speed, and has the advantage of a compressed network structure that facilitates deployment of the model on resource-limited devices.

Beyond Words

Beyond Words

PNeSM: Arbitrary 3D Scene Stylization via Prompt-Based Neural Style Mapping

3D scene stylization refers to transform the appearance of a 3D scene to match a given style image, ensuring that images rendered from different viewpoints exhibit the same style as the given style image, while maintaining the 3D consistency of the stylized scene. Several existing methods have obtained impressive results in stylizing 3D scenes. However, the mod- els proposed by these methods need to be re-trained when applied to a new scene. In other words, their models are cou- pled with a specific scene and cannot adapt to arbitrary other scenes. To address this issue, we propose a novel 3D scene stylization framework to transfer an arbitrary style to an ar- bitrary scene, without any style-related or scene-related re- training. Concretely, we first map the appearance of the 3D scene into a 2D style pattern space, which realizes complete disentanglement of the geometry and appearance of the 3D scene and makes our model be generalized to arbitrary 3D scenes. Then we stylize the appearance of the 3D scene in the 2D style pattern space via a prompt-based 2D stylization al- gorithm. Experimental results demonstrate that our proposed framework is superior to SOTA methods in both visual qual- ity and generalization.

EDA: Evolving and Distinct Anchors for Multimodal Motion Prediction

Motion prediction is a crucial task in autonomous driving, and one of its major challenges lands in the multimodality of future behaviors. Many successful works have utilized mixture models which require identification of positive mixture components, and correspondingly fall into two main lines: prediction-based and anchor-based matching. The prediction clustering phenomenon in prediction-based matching makes it difficult to pick representative trajectories for downstream tasks, while the anchor-based matching suffers from a limited regression capability. In this paper, we introduce a novel paradigm, named Evolving and Distinct Anchors (EDA), to define the positive and negative components for multimodal motion prediction based on mixture models. We enable anchors to evolve and redistribute themselves under specific scenes for an enlarged regression capacity. Furthermore, we select distinct anchors before matching them with the ground truth, which results in impressive scoring performance. Our approach enhances all metrics compared to the baseline MTR, particularly with a notable relative reduction of 13.5% in Miss Rate, resulting in state-of-the-art performance on the Waymo Open Motion Dataset. Appendix and code are available at https://github.com/Longzhong-Lin/EDA.

GSN: Generalisable Segmentation in Neural Radiance Field

Traditional Radiance Field (RF) representations capture details of a specific scene and must be trained afresh on each scene. Semantic feature fields have been added to RFs to facilitate several segmentation tasks. Generalised RF representations learn the principles of view interpolation. A generalised RF can render new views of an unknown and untrained scene, given a few views. We present a way to distil feature fields into the generalised GNT representation. Our GSN representation generates new views of unseen scenes on the fly along with consistent, per-pixel semantic features. This enables multi-view segmentation of arbitrary new scenes. We show different semantic features being distilled into generalised RFs. Our multi-view segmentation results are on par with methods that use traditional RFs. GSN closes the gap between standard and generalisable RF methods significantly. Project Page: https://vinayak-vg.github.io/GSN/