Camera Perspective Research Articles

Camera perspective and skin color: Biased reactions to viral body worn camera videos of police violence

ABSTRACT Body-worn camera and citizen device videos capturing police use-of-force are shared and commented upon widely within social media. This study investigated how point-of-view (POV: onlooker vs. officer perspective) and citizen skin color (dark skin vs. light skin), interacted to affect emotional responses, likelihood to comment and share, and comment on content. A predominantly White sample watched police use-of-force videos in which citizen skin color and camera POV varied. Body-worn camera (BWC) videos in which light-skinned citizens were harmed elicited the most likelihood to comment and share. Further, experienced negative emotion fully mediated this relationship. BWC videos in which dark-skinned citizens were harmed elicited the least negative emotion, the least likelihood to comment, and less normative commentary about officer behaviors.

Visual Parking Occupancy Detection Using Extended Contextual Image Information via a Multi-Branch Output ConvNeXt Network.

Along with society's development, transportation has become a key factor in human daily life, increasing the number of vehicles on the streets. Consequently, the task of finding free parking slots in metropolitan areas can be dramatically challenging, increasing the chance of getting involved in an accident and the carbon footprint, and negatively affecting the driver's health. Therefore, technological resources to deal with parking management and real-time monitoring have become key players in this scenario to speed up the parking process in urban areas. This work proposes a new computer-vision-based system that detects vacant parking spaces in challenging situations using color imagery processed by a novel deep-learning algorithm. This is based on a multi-branch output neural network that maximizes the contextual image information to infer the occupancy of every parking space. Every output infers the occupancy of a specific parking slot using all the input image information, unlike existing approaches, which only use a neighborhood around every slot. This allows it to be very robust to changing illumination conditions, different camera perspectives, and mutual occlusions between parked cars. An extensive evaluation has been performed using several public datasets, proving that the proposed system outperforms existing approaches.

Dynamically occluded action recognition by pigeons.

Identifying the behaviors of organisms is essential for an animal's survival. This ability is particularly challenged when the "actors" are dynamically occluded by other objects and become fragmented as they move through an environment. Even when fragmented in time and across space, humans readily recognize the behavior of these dynamically occluded objects and actors. How animals process such fragmented information, especially when involving motion, remains uncertain. In three experiments, we investigated the ability of six pigeons to discriminate between the running and walking actions of digital animal models when dynamically occluded. The pigeons were tested in a go/no-go procedure using three models that transited behind multiple occluders in a semirealistic scene. Without ever seeing the entirety of the animal model at one time, all the pigeons learned to discriminate among these two behaviors. This discrimination transferred to an unfamiliar model, transit direction, transiting rates, camera perspectives, and occluders. Tests with different static and dynamic features indicated that the pigeons relied on motion features for the discrimination, especially articulated motion. These experiments demonstrate that pigeons, like humans, can discriminate actions even when their view of the actor is fragmented in time and space.

Locate, Tell, and Guide: Enabling Public Cameras to Navigate the Public

Indoor navigation is essential to a wide spectrum of applications in the era of mobile computing. Existing vision-based technologies suffer from both start-up costs and the absence of semantic information for navigation. We observe an opportunity to leverage pervasively deployed surveillance cameras to deal with the above drawbacks and revisit the problem of indoor navigation with a fresh perspective. In this paper, we propose <i>iSAT</i> , a system that enables public surveillance cameras, as indoor navigating satellites, to locate users on the floorplan, tell users with semantic information about the surrounding environment, and guide users with navigation instructions. However, enabling public cameras to navigate is non-trivial due to 3 factors: absence of real scale, disparity of camera perspective, and lack of semantic information. To overcome these challenges, <i>iSAT</i> leverages POI-assisted framework and adopts a novel coordinate transformation algorithm to associate public and mobile cameras, and further attaches semantic information to user location. Extensive experiments in 4 different scenarios show that <i>iSAT</i> achieves a localization accuracy of 0.48m and a navigation success rate of 90.5 percent, outperforming the state-of-th-art systems by <inline-formula><tex-math notation="LaTeX">$&gt; 30\%$</tex-math></inline-formula> . Benefiting from our solution, all areas with public cameras can upgrade to smart spaces with visual navigation services.

Pigeons discount continuously changing perspective during action recognition.

An important challenge for animal and artificial visual systems is separating the system's own motions from the movements of other animals or events. To examine this issue in birds, we conducted three experiments testing four pigeons in a go/no-go action discrimination. The pigeons discriminated whether a digital human model was exhibiting an extended series of articulated motions or one of a set of static poses from the same video. They were required to do so while the rendering camera's perspective changed continually during each trial's 20-s video presentation. Experiment 1 found that pigeons easily discount the camera's continuous motion. Experiments 2 and 3, by testing novel sequences of the behavior, novel behaviors, silhouettes, and a form of conditional discrimination, revealed this to be a general capacity. Overall, the discrimination was predominantly mediated by global action cues, although a small contribution of image-based statistical features was detected. Collectively, the experiments reveal pigeons can readily separate and discount constantly changing perspectives while processing others' actions. (PsycInfo Database Record (c) 2023 APA, all rights reserved).

Linking Thermal Images with 3D Models for FFF Printing

The thermal profile plays a major role in additive manufacturing. Thermal cameras are thus increasingly used for quality monitoring. So far, either full thermal images or metrics extracted from them are used to monitor the manufacturing quality or detect defects. To additionally allow the detection of local anomalies, it is necessary to link the thermal image to the 3D object geometry. We propose a framework that includes steps for filtering object points that are visible from the camera perspective, projecting 3D points onto thermal images and removing pixels that represent the printhead. Our framework can be used for process monitoring and subsequent on-line defect detection which are necessary components for production automation and Industry 4.0 applications. In a validation experiment, we show that the temperature extracted from thermal images and assigned to 1mm × 1mm × 1mm voxels is highly correlated to the temperature measured with type K thermocouples.

Semi-Supervised Bolt Anomaly Detection Based on Local Feature Reconstruction

During the process of high-speed railway online maintenance, anomaly detection of targets in grease injection facility is extremely challenging. Environmental uncertainty (noise disturbance, various illumination conditions, changeable camera perspective, etc.) limits the performance of existing unsupervised anomaly detection methods, which tend to struggle in such complex backgrounds without prior knowledge of anomalies. Therefore, we propose a novel semi-supervised anomaly detection algorithm that only learns from anomaly-free samples and annotations while can recognize both anomaly-free and anomaly targets (key-areas) and detecting anomalous targets simultaneously. We firstly propose a key-area recognition module (KRM) to segment anomaly-free targets and predict coordinates of both anomaly-free and anomalous key-areas’ center point (key-points). By using the predicted coordinates, we generated a circular mask to erase features within the key-areas. Based on the erased features, a local feature reconstruction and segmentation module (LFRSM) is proposed to restore the latent embeddings within the key-areas and segment both anomaly-free and anomalous key-areas. The final anomaly results are obtained by analyzing the original segmentation results of KRM and the reconstructed segmentation results of LFRSM. To validate the effectiveness of the proposed algorithm, we develop a dataset, called BOLT-DET, for bolt anomaly detection in complex scenes. The proposed algorithm outperforms other existing state-of-the-art (SOTA) anomaly detection algorithms (SPADE, DFR, CFLOW, Student-Teacher, PatchCore, and DRAEM) with mIoU of 0.66, AUROC of 0.92 and F1-Measure of 0.69 on BOLT-DET. It can accurately detect anomalies and meet the requirement of practical industrial applications.

Bolt loosening angle detection based on binocular vision

Bolt looseness detection is critical in preventing bolt connection failure. Compared to traditional sensor-based bolt looseness detection, image-based methods are low-cost and contactless and have thus become the highlight of research. However, current monocular vision-based detection methods are prone to error scaused by the camera perspective . In this paper, we present a novel bolt loosening angle detection method based on binocular vision. Key points on the bolt are detected and matched by SuperPoint Gauss network for 3D coordinates reconstruction and motion tracking. The bolt loosening angle is solved by fitting the rotation equation using random sample consensus. Experiments verify the proposed method performs well under different perspectives of camera and illumination conditions with an average error of 1.5°. Comparative test shows our method is superior to the monocular vision-based method in terms of accuracy when there is a large perspective angle. The proposed method is mark-free and robust to various working conditions, which makes it of great value for engineering application.

Joint Cross-Consistency Learning and Multi-Feature Fusion for Person Re-Identification

To solve the problem of inadequate feature extraction by the model due to factors such as occlusion and illumination in person re-identification tasks, this paper proposed a model with a joint cross-consistency learning and multi-feature fusion person re-identification. The attention mechanism and the mixed pooling module were first embedded in the residual network so that the model adaptively focuses on the more valid information in the person images. Secondly, the dataset was randomly divided into two categories according to the camera perspective, and a feature classifier was trained for the two types of datasets respectively. Then, two classifiers with specific knowledge were used to guide the model to extract features unrelated to the camera perspective for the two types of datasets so that the obtained image features were endowed with domain invariance by the model, and the differences in the perspective, attitude, background, and other related information of different images were alleviated. Then, the multi-level features were fused through the feature pyramid to concern the more critical information of the image. Finally, a combination of Cosine Softmax loss, triplet loss, and cluster center loss was proposed to train the model to address the differences of multiple losses in the optimization space. The first accuracy of the proposed model reached 95.9% and 89.7% on the datasets Market-1501 and DukeMTMC-reID, respectively. The results indicated that the proposed model has good feature extraction capability.

Shaping film: A quantitative formal analysis of contemporary empathy-eliciting Hollywood cinema.

Previous research suggests that particular formal features of film, such as the use of close-ups, can affect the levels of empathy experienced by viewers. As empathy is a key aspect of audience’s filmic experience, creative decisions in editing and cinematography may be motivated by the filmmaker’s intention of eliciting empathy. The goal of this study is to investigate what film scenes intended to elicit empathy look like in terms of those visual formal features theoretically or empirically linked to viewer empathy, and if these features converge on something that might be dubbed an empathic style of cinema. Formal features included concern shot scale, face depiction, cut rate, camera perspective and angle, saturation, lighting, motion, and background clutter. Exploratory quantitative formal analyses of scenes sampled from contemporary popular empathy-eliciting Hollywood films (N = 100) revealed that such scenes are, at first glance, highly dissimilar in form. Further investigation through principal component analysis and correlational analysis, however, hints not so much at a singular empathic style of cinema as it does at certain general principles, namely the reduction of perceived distance through close-ups and face depiction, the balancing of arousing features with comprehensible levels of visual complexity, and the prioritization of coherence and reduced visual contrast to enable a smooth viewing experience.

Identity-Unrelated Information Decoupling Model for Vehicle Re-Identification

As an indispensable part of intelligent transportation system (ITS), vehicle re-identification (Re-ID) aims to retrieve all target vehicle images captured from non-overlapping cameras. However, this task remains very challenging due to the variation of camera perspective and the similar appearance among the vehicles with the same type and color, i.e., large intra-class variances and small inter-class variances. Previous methods have made a great progress on vehicle Re-ID by leveraging local details and aligning local features, this issue is still far from being solved. In this work, we propose to decouple identity-unrelated information from vehicle representation, tackling the problems of camera perspective variation and vehicle appearance similarity. The keypoint of this method is to learn a distinguishable feature embedding that is independent of identity-unrelated information. Specially, the novel Identity-Unrelated Information Decoupling (IUID) paradigm is designed to learn invariant features of the vehicle with the same ID in different scenes. In our approach, identity-unrelated information can be divided into two kinds of information, i.e., camera perspective information and background information. For the former, through a feature-level camera generative adversarial module, we can decouple camera perspective information from the feature embedding after extracting invariant features across different cameras perspective. For the latter, we propose a vehicle-mask transformer to enhance the attention of the model on local details while reducing the impact of background information. Extensive experiments on two public datasets demonstrate the superiority of IUID over the current state-of-the-arts methods.

Swedish Cold War history on YouTube – committed amateurs and heritagization from below

ABSTRACT This study explores the meaning-making of amateur videos on YouTube pertaining to the Swedish Cold War heritage and it contributes with a discussion on how videographic conventions and social media platform logics intervene in the ongoing informal heritagization of the Cold War era. The heritagization process of the Cold War remains in Sweden during the 1990s and the beginning of the new millennium coincided with the advent of the online society. The process seemed to resonate of the democratic ideals from the discourse of Heritage from below. Now it seemed like anyone had the possibility to become a heritage producer. However, heritagization from below came with unintended implications. The analysis of YouTube videos in this study suggests that the vernacular Cold War heritage is colored by an easily digested format containing of moving still pictures, mood-inducing soundtracks, luring camera perspectives, rhythmic editing, and genre loans from video games and horror films, which tend to safeguard the naturalness of filmed sites and an entire era.

VivaSight Single-Lumen Tube Combined With Hyperangulated Videolaryngoscopy to Rescue Failed Tracheal Intubation in a Patient With Goldenhar Syndrome: A Case Report

This report describes a patient with Goldenhar syndrome undergoing anesthesia for whom Macintosh videolaryngoscopy failed, as the epiglottis was adhered to the posterior pharynx and could not be lifted with a tracheal introducer (Cormack-Lehane grade 3B). Hyperangulated videolaryngoscopy revealed only the arytenoids (Cormack-Lehane grade 2B), even after direct lifting of the epiglottis, and endotracheal tube advancement failed due to unclear tissue resistance. Hyperangulated videolaryngoscopy was combined with a tube-mounted camera (VivaSight single lumen tube). The combination of both camera perspectives was successfully used to allow placement of the endotracheal tube underneath the epiglottis and through the vocal cords.

Race, Melancholia, Midsommar

Race, Melancholia, Midsommar Monica L. Miller (bio) I hardly know where to look, where to rest my gaze, whom to follow, and how to do so in this blinding sunlight. I've been here before; this is like my first midsommar (summer solstice), in what I thought was a fantasy of Sweden but turned out in some ways to be real or at least a curious affekt (affect) of the people and the place. I remember being in a field adjacent to a series of red summerhouses, everyone was dressed in white, their summer finest. We sang and drank and sang again during a lunch of pickled herring, smoked salmon, fresh potatoes, elderflower juice, and Swedish snaps, all served family style on long tables decorated with fresh flowers. The herring and snaps were new to me, the fish both slimy and oily despite the delicious mustard sauce and the snaps so full of alcohol that they combined purification and inebriation. We ran to the maypole and danced, pretending to be, as I learned later, little frogs and pigs. The blinding light softened to a beautiful glow as the sun did not set and we kept dancing, drinking, and eating strawberries and whipped cream. I was far away from home, a stranger in a strange land, celebrating midsommar with my Swedish partner and his friends. It was beautiful and strange. In Ari Aster's film Midsommar when the car carrying Dani, Christian, Josh, Mark, and Pelle drives into the Swedish countryside toward Hårga and the camera's perspective slowly, gently, definitively flips upside down, I feel a sharp pang of (mis)recognition, of having felt on the precipice of a similar summer scene. And yet as I stare at the screen at the [End Page 362] beginning of this fictional filmic midsommar celebration, I hardly know where to look because I am (over)identifying with nearly everyone, American and Swede alike. I don't know where to look when the Americans arrive in Hårga because all I see is white—white people in white clothing, all in blinding sunlight—so I decide to fixate on Josh. Like me, he's African American and a scholar, continually analyzing and taking notes; he's traveled to Sweden to have fun and to do fieldwork for his thesis. His is a scenario that is familiar to me. Though I am not an anthropologist like Josh, I am a literary and cultural historian; more importantly, I've seen enough horror films to know that people like us, Black and too curious, will likely not survive these two and a half hours, even and maybe especially in Sweden. I wonder how he will navigate this situation, where he will end up, what his presence and the manner and timing of his death can or will tell us about how to read race in the fictional place of this film or Sweden, as metaphor. I'm thinking about affekt again and belonging and remembering that at the end of Midsommar during the ritual burning of the nine offerings, the Hårgans are said to purge oheliga affekter (unholy affects) out of their society. I'm thinking also about the fact that Sweden is officially and in public discourse a "place without race" with a strong sense of itself as a color-blind social democracy that is exceptional in Europe because it had little to no colonial enterprise or overseas empire. Historically, the homogeny of the population and the overwhelming nature of whiteness as a sociocultural norm has meant that instead of identifying as "people of color," racialized Swedes have identified as icke-vit (nonwhite) instead of, for example, in the case of people of African descent, Black or Afro-Swedish, terms of self-identification more in use today. I wonder if or how a place for Blackness will or can be made in this film. When thinking about how to apprehend difference and Black people at midsommar celebrations or in Sweden more generally, I think of my own experiences, lovely long days that became even richer as I learned the language, songs, and traditions over the years. Early on I bonded with the other foreigners...

Learners' Responses to a Virtual Cadaver Dissection Nerve Course in the COVID Era: A Survey Study.

Background Virtual education is an evolving method for teaching medical learners. During the coronavirus disease 2019 pandemic, remote learning has provided a replacement for conferences, lectures, and meetings, but has not been described as a method for conducting a cadaver dissection. We aim to demonstrate how learners perceive a virtual cadaver dissection as an alternative to live dissection. Methods A virtual cadaver dissection was performed to demonstrate several upper extremity nerve procedures. These procedures were livestreamed as part of an educational event with multimedia and interactive audience questions. Participants were queried both during and after the session regarding their perceptions of this teaching modality. Results Attendance of a virtual dissection held for three plastic surgery training institutions began at 100 and finished with 70 participants. Intrasession response rates from the audience varied between 68 and 75%, of which 75% strongly agreed that they were satisfied with the virtual environment. The audience strongly agreed or agreed that the addition of multimedia captions (88%), magnified video loupe views (82%), and split-screen multicast view (64%) was beneficial. Postsession response rate was 27%, and generally reflected a positive perspective about the content of the session. Conclusions Virtual cadaver dissection is an effective modality for teaching surgical procedures and can be enhanced through technologies such as video loupes and multiple camera perspectives. The audience viewed the virtual cadaver dissection as a beneficial adjunct to surgical education. This format may also make in-person cadaver courses more effective by improving visualization and allowing for anatomic references to be displayed synchronously.

Unmanned Aerial Vehicles: A Literature Review

In recent years, Unmanned Aerial Vehicles (UAVs) have grown and increased in applications because of computational simplicity and adaptive control capacity with strong support from both civilian and military sectors. The applications of UAVs in various military, commercial and civilian areas have led to sustainable results. The application areas include but are not limited to oil &amp; gas, cargo transport, geographic mapping, aerial photography, health care, and disaster management. The success of the UAV application missions is completely dependent on the accuracy in control provided by the flight controllers. Thus, there is a need for accurate, robust, and adaptive flight controllers. UAV dynamics modeling and identification and control of these vehicles are still major active areas of research and development. They pose severe challenges due to the vehicle's complex design, inherently nonlinear, and time-varying dynamics. The main goal of this paper is to identify the past research trends and recent improvements in UAVs. Furthermore, this paper discusses a comprehensive literature review according to the optimized objectives, solution techniques, and applications of UAVs such as Cargo Transport, Disaster Management, etc. According to the literature review, aerial photography is one of the applications of smart UAVs. The reliability of image matching across multiple camera perspectives, angles, and positions encourages computer vision approaches for UAV navigation, opening the way for future researchers to develop vision applications. This article presents a comprehensive literature review discussing the importance of UAV applications related to cost-effectiveness and versatility. Furthermore, a detailed survey of system modeling identification and control techniques is presented.

(Always) Playing the Camera: Cyborg Vision and Embodied Surveillance in Digital Games

As the increasingly ubiquitous field of surveillance has transformed how we interact with each other and the world around us, surveillance interactions with virtual others in virtual worlds have gone largely unnoticed. This article examines representations of digital games’ diegetic surveillance cameras and their relation to the player character and player. Building on a dataset of forty-one titles and in-depth analyses of two 2020 digital games that present embodied surveillance camera perspectives, Final Fantasy VII Remake (Square Enix 2020) and Watch Dogs: Legion (Ubisoft Toronto 2020), I demonstrate that the camera is crucial in how we organize, understand, and maneuver the fictional environment and its inhabitants. These digital games reveal how both surveillance power fantasies and their critique can coexist within a space of play. Moreover, digital games often present a perspective that blurs the boundaries between the physical and the technically mediated through a flattening of the player’s “camera” screen and in-game surveillance cameras. Embodied surveillance cameras in digital games make the camera metaphor explicit as an aesthetic, narrative, and mechanical preoccupation. We think and play with and through cameras, drawing attention to and problematizing the partial perspectives with which worlds are viewed. I propose the term cyborg vision to account for this simultaneously human and nonhuman vision that’s both pluralistic and situated and argue that, through cyborg vision, digital games offer an embodied experience of surveillance that’s going to be increasingly relevant in the future.

#socialiseresponsibly. Analyzing the Rhetorical Structure of Heineken TV Commercials During the Pandemic

This work analyzes the efforts by the company Heineken to address in their marketing campaigns the social responsibility and wellbeing of their customers during the pandemic. By particularly looking at the audio-visual design strategies used in 15 TV commercials produced and published during the years 2020–2022, a multimodal analysis of the filmic montage structures examines how the company establishes a coherent process of core marketing and convinces recipients to socialize responsibly under pandemic conditions. The analysis follows the current trend in empirical multimodality research to pursue analyses of digital corpora with the theories and methods developed in the field. Particularly, it uses an ELAN annotation scheme for the description of specific camera techniques (such as camera distance and perspective) as well as the discourse relations holding between the shots and segments identified in the commercials. The findings show that the company mainly uses one specific type of rhetorical structure, the narrative commentary, that puts the social actors shown in many different scenes into the foreground and therefore strongly focuses on the company's strategy of positive social behavior.

What do you see? Understanding perceptions of police use of force videos as a function of the camera perspective.

Some research suggests that video-recorded police incidents may be subject to a camera perspective bias. This study examined whether the camera angle of a recorded police use of force encounter influenced interpretation of the video. Participants (n=330) viewed a video-recorded simulated use of force scenario in one of four camera angle conditions (body worn camera, bystander camera, security camera, all three camera angles), and then rated the conduct of the police officer and the subject. Participants' attitudes towards the police and legal system were also examined. Results indicated that camera angle did not directly impact viewers' judgment of the scenario, but pre-existing biases about the police guided their interpretations of certain camera angles. Importantly, however, this was not the case for those who viewed the body worn camera angle. These results help us understand the implications of relying on video recordings of police incidents.

A Survey on GAN-Based Data Augmentation for Hand Pose Estimation Problem

Deep learning solutions for hand pose estimation are now very reliant on comprehensive datasets covering diverse camera perspectives, lighting conditions, shapes, and pose variations. While acquiring such datasets is a challenging task, several studies circumvent this problem by exploiting synthetic data, but this does not guarantee that they will work well in real situations mainly due to the gap between the distribution of synthetic and real data. One recent popular solution to the domain shift problem is learning the mapping function between different domains through generative adversarial networks. In this study, we present a comprehensive study on effective hand pose estimation approaches, which are comprised of the leveraged generative adversarial network (GAN), providing a comprehensive training dataset with different modalities. Benefiting from GAN, these algorithms can augment data to a variety of hand shapes and poses where data manipulation is intuitively controlled and greatly realistic. Next, we present related hand pose datasets and performance comparison of some of these methods for the hand pose estimation problem. The quantitative and qualitative results indicate that the state-of-the-art hand pose estimators can be greatly improved with the aid of the training data generated by these GAN-based data augmentation methods. These methods are able to beat the baseline approaches with better visual quality and higher values in most of the metrics (PCK and ME) on both the STB and NYU datasets. Finally, in conclusion, the limitation of the current methods and future directions are discussed.