Object Categories Research Articles

Joint Optimization-Based Texture and Geometry Enhancement Method for Single-Image-Based 3D Content Creation

Recent studies have explored the generation of three-dimensional (3D) meshes from single images. A key challenge in this area is the difficulty of improving both the generalization and detail simultaneously in 3D mesh generation. To address this issue, existing methods utilize fixed-resolution mesh features to train networks for generalization. This approach is capable of generating the overall 3D shape without limitations on object categories. However, the generated shape often exhibits a blurred surface and suffers from suboptimal texture resolution due to the fixed-resolution mesh features. In this study, we propose a joint optimization method that enhances geometry and texture by integrating generalized 3D mesh generation with adjustable mesh resolution. Specifically, we apply an inverse-rendering-based remeshing technique that enables the estimation of complex-shaped mesh estimations without relying on fixed-resolution structures. After remeshing, we enhance the texture to improve the detailed quality of the remeshed mesh via a texture enhancement diffusion model. By separating the tasks of generalization, detailed geometry estimation, and texture enhancement and adapting different target features for each specific network, the proposed joint optimization method effectively addresses the characteristics of individual objects, resulting in increased surface detail and the generation of high-quality textures. Experimental results on the Google Scanned Objects and ShapeNet datasets demonstrate that the proposed method significantly improves the accuracy of 3D geometry and texture estimation, as evaluated by the PSNR, SSIM, LPIPS, and CD metrics.

What do we see behind an occluder? Amodal completion of statistical properties in complex objects.

When a spiky object is occluded, we expect its spiky features tocontinue behind the occluder. Although many real-world objects contain complex features, it is unclear how more complex features are amodally completed and whether this process is automatic. To investigate this issue, we created pairs of displays with identical contour edges up to the point of occlusion, but with occluded portions exchanged. We then asked participants to search for oddball targets among distractors and asked whether relations between searches involving occluded displays would match better with relations between searches involving completions that are either globally consistent or inconsistent with the visible portions of these displays. Across two experiments involving simple and complex shapes, search times involving occluded displays matched better with those involving globally consistent compared with inconsistent displays. Analogous analyses on deep networks pretrained for object categorization revealed a similar pattern of results for simple but not complex shapes. Thus, deep networks seem to extrapolate simple occluded contours but not more complex contours. Taken together, our results show that amodal completion in humans is sophisticated and can be based on extrapolating global statistical properties.

PCLC‐Net: Point Cloud Completion in Arbitrary Poses with Learnable Canonical Space

AbstractRecovering the complete structure from partial point clouds in arbitrary poses is challenging. Recently, many efforts have been made to address this problem by developing SO(3)‐equivariant completion networks or aligning the partial point clouds with a predefined canonical space before completion. However, these approaches are limited to random rotations only or demand costly pose annotation for model training. In this paper, we present a novel Network for Point cloud Completion with Learnable Canonical space (PCLC‐Net) to reduce the need for pose annotations and extract SE(3)‐invariant geometry features to improve the completion quality in arbitrary poses. Without pose annotations, our PCLC‐Net utilizes self‐supervised pose estimation to align the input partial point clouds to a canonical space that is learnable for an object category and subsequently performs shape completion in the learned canonical space. Our PCLC‐Net can complete partial point clouds with arbitrary SE(3) poses without requiring pose annotations for supervision. Our PCLC‐Net achieves state‐of‐the‐art results on shape completion with arbitrary SE(3) poses on both synthetic and real scanned data. To the best of our knowledge, our method is the first to achieve shape completion in arbitrary poses without pose annotations during network training.

Skin feature point tracking using deep feature encodings

AbstractFacial feature tracking is a key component of imaging ballistocardiography (BCG) where accurate quantification of the displacement of facial keypoints is needed for good heart rate estimation. Skin feature tracking enables video-based quantification of motor degradation in Parkinson’s disease. While traditional computer vision algorithms like Scale Invariant Feature Transform (SIFT), Speeded-Up Robust Features (SURF), and Lucas-Kanade method (LK) have been benchmarks due to their efficiency and accuracy, they often struggle with challenges like affine transformations and changes in illumination. In response, we propose a pipeline for feature tracking, that applies a convolutional stacked autoencoder to identify the most similar crop in an image to a reference crop containing the feature of interest. The autoencoder learns to represent image crops into deep feature encodings specific to the object category it is trained upon. We train the autoencoder on facial images and validate its ability to track skin features in general using manually labelled face and hand videos of small and large motion recorded in our lab. Our evaluation protocol is comprehensive, including quantification of errors in human annotation. The tracking errors of distinctive skin features (moles) are so small that we cannot exclude the fact that they stem from the manual labelling based on a $$\chi ^2$$ χ 2 -test. With a mean error of 0.6–3.3 pixels, our method outperformed the other methods in all but one scenario. More importantly, our method was the only one that did not diverge. We also compare our method with the latest state-of-the-art transformer for feature matching by Google—Omnimotion. Our results indicate that our method is superior at tracking different skin features under large motion conditions and that it creates better feature descriptors for tracking, matching, and image registration compared to both traditional algorithms and the latest Omnimotion.

PHÂN TÍCH THỰC TRẠNG VÀ SỰ KHÁC BIỆT CỦA CÁC NGUYÊN NHÂN LÃNG PHÍ THỜI GIAN VỚI KẾT QUẢ HỌC TẬP CỦA SINH VIÊN

To address the causes of wasting time that have or do not have an impact on students' academic achievement, this study was conducted utilizing quantitative research methods, alongside a variety of research methods, such as document research, an exploratory survey of 307 students, an official survey with 6,670 statistically valid student’s responses, interviews, exploratory factor analysis (EFA), scale reliability analysis, descriptive statistics, and inferential statistics. Consequently, all of the 11 causes of wasting time that belong to the two categories of objective and subjective causes are at the influential level. Whether in the group of objective or subjective causes, reasons related to the use of technology are the leading cause of wasting students' time. In addition, the study also found that there is a significant difference between student's academic results, with the main trend of students with fair academic performance having more wasted time when compared to groups of students with good and excellent grades. Therefore, to have good learning results, students need to study and practice time management skills to increase the efficiency of time use and avoid time-consuming factors.

Using fMRI to examine nonlinear mixed selectivity tuning to task and category in the human brain

Abstract Recent experimental and theoretical work has shown that nonlinear mixed selectivity, where neurons exhibit interaction effects in their tuning to multiple variables (e.g., stimulus and task) plays a key role in enabling the primate brain to form representations that can adapt to changing task contexts. Thus far all such studies have relied on invasive neural recording techniques. In this study, we demonstrate the feasibility of measuring nonlinear mixed selectivity tuning in the human brain noninvasively using fMRI pattern decoding. To do so, we examined the joint representation of object category and task information across human early, ventral stream, and dorsal stream areas while participants performed either an oddball detection task or a one-back repetition detection task on the same stimuli. These tasks were chosen to equate spatial, object-based, and feature-based attention, in order to test whether task modulations of visual representations still occur when the inputs to visual processing are kept constant between the two tasks, with only the subsequent cognitive operations varying. We found moderate but significant evidence for nonlinear mixed selectivity tuning to object category and task in fMRI response patterns in both human ventral and dorsal areas, suggesting that neurons exhibiting nonlinear mixed selectivity for category and task not only exist in these regions, but cluster at a scale visible to fMRI. Importantly, while such coding in ventral areas corresponds to a rotation or shift in the object representational geometry without changing the representational content (i.e., with the relative similarity among the categories preserved), nonlinear mixed selectivity coding in dorsal areas corresponds to a reshaping of representational geometry, indicative of a change in representational content.

DroneSilient (drone + resilient): an anti-drone system

AbstractIt is imperative to take a holistic strategy to thwarting drone threats, including the identification of drones and drone-like gadgets like ornithopters that visually imitate birds. In this study, we present the DroneSilient System, a novel anti-drone system that combines different parts. The DroneSilient system includes components that connect to RF identification technology and image-capture technology. A modified bloom filter method is used to further identify the recognized object after a drone-like object has been found, allowing for the differentiation between regular drones, ornithopters, and genuine birds. The CNN (Convolutional Neural Network) method, created using the Google Cloud Platform and AutoML widget, is used in our model for object identification and categorization. DroneSilient has an RF sensor that can identify and imitate the threat presented by recognized drones. Convolutional Network, Modified Blooms Algorithm, RFID, and RF Sensor systems are all integrated into the DroneSilient system as part of this methodology combination, which provides a thorough method for identifying and eliminating drone threats. The bloom filter proposed takes 27.6 to 12.4 microseconds. Overall time for handling the unauthorized drone will be less than 180 s. By tackling every facet of the problem, our strategy outperforms many current anti-drone solutions in terms of functionality.

Prior probability biases perceptual choices by modulating the accumulation rate, rather than the baseline, of decision evidence

Abstract The prior probability of an upcoming stimulus has been shown to influence the formation of perceptual decisions. Computationally these effects have typically been attributed to changes in the starting point (i.e. baseline) of evidence accumulation in sequential sampling models. More recently, it has also been proposed that prior probability might additionally lead to changes in the rate of evidence accumulation. Here, we introduce a neurally-informed behavioural modelling approach to understand whether prior probability influences the starting point, the rate of evidence accumulation or both. To this end, we employ a well-established visual object categorization task for which two neural components underpinning participants’ choices have been characterised using single-trial analysis of the electroencephalogram. These components are reliable measures of trial-by-trial variability in the quality of the relevant decision evidence, which we use to constrain the estimation of a hierarchical drift diffusion model of perceptual choice. We find that, unlike previous computational accounts, constraining the model with the endogenous variability in the relevant decision evidence, results in prior probability effects being explained primarily by changes in the rate of evidence accumulation rather than changes in the starting point or a combination of both. Ultimately, our neurally-informed modelling approach helps disambiguate the mechanistic effect of prior probability on perceptual decision formation, suggesting that prior probability biases primarily the interpretation of sensory evidence towards the most likely stimulus.

A Robust TrafficSignNet Algorithm for Enhanced Traffic Sign Recognition in Autonomous Vehicles Under Varying Light Conditions

Recent years have witnessed significant advancements in machine perception, particularly in the context of self-driving vehicles. The accurate detection and interpretation of road signs by these vehicles are crucial for enhancing safety, intelligence, and efficiency on the roads. Consequently, there is a growing body of research dedicated to improving traffic sign recognition technologies, a key component of intelligent transportation systems. Annual statistics highlight numerous accidents attributable to factors such as excessive speed, variable lighting conditions, and the misinterpretation of traffic signs. In response to these challenges, a novel approach for the rapid and reliable recognition of traffic signs by moving vehicles has been developed. This approach leverages a custom dataset encompassing twelve object categories and seven subcategories, reflective of road sign diversities encountered in India. A specialized algorithm, TrafficSignNet, was devised to specifically identify signs related to speed, turning, zones, and bumps. This algorithm was trained on a comprehensive dataset comprising 4,962 images, with its performance evaluated using 705 images from real traffic scenarios. The evaluation demonstrates that the model achieves remarkable accuracy across various lighting conditions, processing up to 12 frames per second. This processing rate is compatible with the high-definition standards of contemporary vehicle cameras, which is 1280 × 720 pixels. The model's effectiveness is quantified through accuracy, precision, recall, and F1 score, with respective values of 0.985, 0.978, 0.964, and 0.971, showcasing its potential to significantly contribute to the advancement of smart transportation systems.

Thalamic Roles in Conscious Perception Revealed by Low-Intensity Focused Ultrasound Neuromodulation.

The neural basis of conscious perception remains incompletely understood. While cortical mechanisms of conscious content have been extensively investigated, the role of subcortical structures, including the thalamus, remains less explored. We aim to elucidate the causal contributions of different thalamic regions to conscious perception using transcranial low-intensity focused ultrasound (LIFU) neuromodulation. We hypothesize that modulating different thalamic regions would result in distinct perceptual outcomes. We apply LIFU in human volunteers to investigate region-specific and sonication parameter-dependent effects. We target anterior (transmodal-dominant) and posterior (unimodal-dominant) thalamic regions, further divided into ventral and dorsal regions, while participants perform a near-threshold visual perception task. Task performance is evaluated using Signal Detection Theory metrics. We find that the high duty cycle stimulation of the ventral anterior thalamus enhanced object recognition sensitivity. We also observe a general (i.e., region-independent) effect of LIFU on decision bias (i.e., a tendency toward a particular response) and object categorization accuracy. Specifically, high duty cycle stimulation decreases categorization accuracy, whereas low duty cycle shifts decision bias towards a more conservative stance. In conclusion, our results provide causal insight into the functional organization of the thalamus in shaping human visual experience and highlight the unique role of the transmodal-dominant ventral anterior thalamus.

Flat comodules and contramodules as directed colimits, and cotorsion periodicity

AbstractThis paper is a follow-up to Positselski and Št’ovíček (Flat quasi-coherent sheaves as directed colimits, and quasi-coherent cotorsion periodicity. Electronic preprint arXiv:2212.09639 [math.AG]). We consider two algebraic settings of comodules over a coring and contramodules over a topological ring with a countable base of two-sided ideals. These correspond to two (noncommutative) algebraic geometry settings of certain kind of stacks and ind-affine ind-schemes. In the context of a coring $${\mathcal {C}}$$ C over a noncommutative ring A, we show that all A-flat $${\mathcal {C}}$$ C -comodules are $$\aleph _1$$ ℵ 1 -directed colimits of A-countably presentable A-flat $${\mathcal {C}}$$ C -comodules. In the context of a complete, separated topological ring $${\mathfrak {R}}$$ R with a countable base of neighborhoods of zero consisting of two-sided ideals, we prove that all flat $${\mathfrak {R}}$$ R -contramodules are $$\aleph _1$$ ℵ 1 -directed colimits of countably presentable flat $${\mathfrak {R}}$$ R -contramodules. We also describe arbitrary complexes, short exact sequences, and pure acyclic complexes of A-flat $${\mathcal {C}}$$ C -comodules and flat $${\mathfrak {R}}$$ R -contramodules as $$\aleph _1$$ ℵ 1 -directed colimits of similar complexes of countably presentable objects. The arguments are based on a very general category-theoretic technique going back to an unpublished 1977 preprint of Ulmer and rediscovered in Positselski (Notes on limits of accessible categories. Electronic preprint arXiv:2310.16773 [math.CT]). Applications to cotorsion periodicity and coderived categories of flat objects in the respective settings are discussed. In particular, in any acyclic complex of cotorsion $${\mathfrak {R}}$$ R -contramodules, all the contramodules of cocycles are cotorsion.

Open Domain Target Detection Based on Image-text Alignment Large Model under Railway Smart Construction Site

Abstract Open domain target detection plays a crucial role in ensuring people’s life safety in Railway Smart Construction Site. This paper proposes a target detection method that is not limited by the number of labels, which is different from traditional models that are usually trained and tested on a fixed set of object categories (or a set of labels). Firstly, the pre-training GLIP model is utilized to train the collected image-text pairing data to generate grounding boxes. Secondly, the pseudo-labeled data formed by associating these boxes with the corresponding text descriptions were combined with the original image-text data to form an enhanced dataset. The final model performs inference based on the prompt and incoming image data to determine the localization information. The GLIP model effectively aligns image and text information and enhances its ability to recognize and handle problems by observing and understanding new situations.

Early perceptual locus of suppression during the attentional blink.

The attentional blink (AB) demonstrates that recognizing the second of two targets (T1 and T2) is difficult when they appear in close succession in a rapid serial visual presentation (RSVP) stream. The AB has been widely accepted as a suppression of T2 processing at the postperceptual stage. The current event-related potential study updates this view by demonstrating the existence of an early perceptual locus of suppression during the AB. Using line drawings of real-life objects as RSVP items, we required participants in Experiment 1 to either discriminate the exact identities or simply classify the object categories of T1 and T2, and in Experiment 2, we instructed participants to discriminate either T1 and T2 identities (dual-target task) or only T2 identity (single-target task) to invalidate the temporal expectation as an alternative account. The results of Experiments 1 and 2 showed that the T2-elicited first positive peak (P1) component was consistently decreased at Lag 3 whenever a dual-target, but not single-target, task was required, and the magnitude of this P1 suppression was significantly predictive of the behavioral AB magnitude in each dual-target task. When the RSVP items were substituted by classic but size-matched alphanumeric characters in Experiment 3, no P1 suppression was evident as expected, ruling out the large stimulus size as an alternative interpretation. These findings provide the strongest evidence to date that the AB can begin to suppress T2 processing at a very early perceptual stage, at least when observers encounter RSVP items of real-life objects, which calls for more flexible cognitive models for the AB. (PsycInfo Database Record (c) 2024 APA, all rights reserved).

A Semantics Analysis of Personification in Percy Bysshe Shelley’s Poetry

This research aims to identify the types of personification and the contextual meaning behind the personification used in Percy Bysshe Shelley’s poetry. This research uses descriptive qualitative method as the purpose of making this research is to analyze the types of personification and the contextual meaning behind the personification used. This research uses semantics studies with several theoretical sources that refer to the theory of Personification from Paxson (1994) and the theory of contextual meaning by Chaer (2009). The results of the research, 41 data contain personification expression was found. There are 12 data of Prosopographia and 29 Prosopopopeia. 12 data of Prosopographia is classified into five sub-classes: Emotions, Characteristic, Concept, State and Time. The research reveals 7 data of Prosopographia in Concept category, 2 data in State category, 3 data in Time category, and 0 data found in Emotions and Characteristic categories. Out of 29 data of Prosopopopeia, the researcher also classified the data into three sub-classes: Living things, Places and Inanimate object. There are 3 data of Prosopopopeia in Living things category, 3 data in Places category and 23 data in Inanimate objects category.

Learning to segment self-generated from externally caused optic flow through sensorimotor mismatch circuits

Efficient sensory detection requires the capacity to ignore task-irrelevant information, for example when optic flow patterns created by egomotion need to be disentangled from object perception. To investigate how this is achieved in the visual system, predictive coding with sensorimotor mismatch detection is an attractive starting point. Indeed, experimental evidence for sensorimotor mismatch signals in early visual areas exists, but it is not understood how they are integrated into cortical networks that perform input segmentation and categorization. Our model advances a biologically plausible solution by extending predictive coding models with the ability to distinguish self-generated from externally caused optic flow. We first show that a simple three neuron circuit produces experience-dependent sensorimotor mismatch responses, in agreement with calcium imaging data from mice. This microcircuit is then integrated into a neural network with two generative streams. The motor-to-visual stream consists of parallel microcircuits between motor and visual areas and learns to spatially predict optic flow resulting from self-motion. The second stream bidirectionally connects a motion-selective higher visual area (mHVA) to V1, assigning a crucial role to the abundant feedback connections to V1: the maintenance of a generative model of externally caused optic flow. In the model, area mHVA learns to segment moving objects from the background, and facilitates object categorization. Based on shared neurocomputational principles across species, the model also maps onto primate visual cortex. Our work extends Hebbian predictive coding to sensorimotor settings, in which the agent actively moves - and learns to predict the consequences of its own movements.

Contrastive learning explains the emergence and function of visual category-selective regions.

Modular and distributed coding theories of category selectivity along the human ventral visual stream have long existed in tension. Here, we present a reconciling framework-contrastive coding-based on a series of analyses relating category selectivity within biological and artificial neural networks. We discover that, in models trained with contrastive self-supervised objectives over a rich natural image diet, category-selective tuning naturally emerges for faces, bodies, scenes, and words. Further, lesions of these model units lead to selective, dissociable recognition deficits, highlighting their distinct functional roles in information processing. Finally, these pre-identified units can predict neural responses in all corresponding face-, scene-, body-, and word-selective regions of human visual cortex, under a highly constrained sparse positive encoding procedure. The success of this single model indicates that brain-like functional specialization can emerge without category-specific learning pressures, as the system learns to untangle rich image content. Contrastive coding, therefore, provides a unifying account of object category emergence and representation in the human brain.

World Heritage: features, composition, and tourist significance

It is proved that the idea of world heritage and its conceptual form of legal imple- mentation created a new institution of world heritage in international law. At the end of the 20th century for the first time, the term «world cultural and natural heritage» became widely used to define those natural and cultural objects that are of outstanding importance not only for the people of a certain country but also for all humanity as a whole. Among the main advantages of granting the status of a world heritage site are the following: additional guarantees of preservation and integrity of unique natural and cultural complexes; increasing the prestige of the territories and institutions that take care of them; promoting the promotion of objects included in the List and the development of alternative types of nature use (primarily, ecological tourism); ensuring priority in attracting financial funds for the support of objects of world cultural and natural heritage, first of all, from the World Heritage Fund; promoting the organization of monitoring and control over the state of preservation of natural and cultural objects. Currently, the World Heritage List consists of three categoriesof objects: cultural heritage, natural phenomena, and cultural-natural (mixed) objects. It is noted that every year the number of world heritage sites is steadily increasing, and by 2020 their number has already reached 1,121 sites, including 869 cultural sites, 213 natural sites, and 39 mixed sites. From 1978 until now, the number of nominated objects has constantly increased. However, in order to achieve representativeness, balance, and credibility of the UNESCO Heritage List, certain restrictions on the number of nominations per year have been introduced since 2018. Now their total number cannot exceed 45 nominations per year and only one nomination from the state. They are distributed very unevenly by regions of the world. More than 75% of the world’s cultural heritage is concentrated in Eurasian countries. The countries of America and Africa hold the leadership in the number of natural monuments in the world. The basis of the interest of the countries of the world in the inclusion of objects in the List of World Heritage is the possibility of obtaining profits from the visit of numerous tourists to this cultural or natural attraction. About 225 historic centers of cities around the world are included in the UNESCO Heritage List, including Paris, London, Istanbul, Vienna, Athens, Florence, Venice, St. Petersburg, Krakow, Riga, Jerusalem, Mecca, Mexico City, Quito, Quebec, Rio de Janeiro -Janeiro, etc. It was found that individual objects of cultural heritage are also very popular (the Great Wall of China, St. Peter’s Basilica in the Vatican, the Colosseum, etc.).

Investigation of Superstitious Behavior Tendencies of Professional Football Players and Coaches

This study aims to elucidate the superstitious behavior tendencies of professional football players and club coaches. The research group consists of 22 football players and 9 football coaches selected through the "Criterion Sampling" method. The research is considered as basic qualitative research. These are qualitative research studies in which data are collected through documents, interviews, or observation analyses. The research design employed in this study is phenomenology (phenomenological research design) which is a qualitative research design. In this research, “Superstitious Belief” has been selected as the phenomenon and the meanings attributed to this phenomenon by football players and coaches are attempted to be revealed. Semi-structured interview forms have been prepared to identify the superstitious behavior tendencies of players and coaches. In this context, open-ended questions are asked to the participants to determine their superstitious behavior tendencies. The data obtained have been analyzed using the content analysis method. According to the findings, the Superstitious Behavior Tendency comprises a theme, a sub-theme, and six categories. Under the sub-theme of superstitious behavior tendencies among professional football players and coaches are the categories of Clothing-Appearance, Lucky Objects, Pre-match Ritual, In-match Behavior, Team Ritual, Faith and Prayer. The opinions of players and coaches regarding the questions prepared for this theme are presented. Word clouds have been created illustrating prominent codes under the identified categories based on data obtained from interviews with participants. In conclusion, beyond effective preparation and training phases for performance enhancement, it is discerned that some superstitious behavior tendencies exist among players, contributing to elevating their motivation and self-confidence to higher levels.

Responses in human early visual cortex are more sensitive to task difficulty than object category.

Responses in human early visual cortex are more sensitive to task difficulty than object category.

Using texture synthesis to identify the features supporting coarse and fine object categorization

Using texture synthesis to identify the features supporting coarse and fine object categorization