In this work, we present AlamViT, an AI model with a novel temporal Swin-enhanced MobileViT architecture for pain recognition from facial expressions. Despite efforts in the field of pain recognition from facial expressions, the field remains immature due to computational costs and the lack of local fine-grained spatial feature extraction in transformerbased models. Moreover, CNN-based models applied in the field lack global context. AlamViT, with its MobileViT-based CNNTransformer architecture, can capture both local and global contexts. Our work further reduces computational costs by applying a windowed Swin attention for the spatial domain. Also, it appends a temporal module to capture the expressions evolving pain. The model performs binary classification (Pain vs. No Pain). We employ two facial-expression video datasets to train and evaluate the model separately: BioVid and AI4PAIN. For BioVid, we use Baseline (BL01) and Pain Level 4 (PA04) facial expression videos. Whereas for AI4PAIN, Baseline/Rest and High Pain clips were used. As per our knowledge, this is the first paper to use Swin-style attention in MobileViT.

Artificial intelligence (AI) has become increasingly popular in recent years, due to advancements in computing power and data availability, as well as its broad range of applications in various industries. ChatGPT is a state-of-the-art language model developed by OpenAI, capable of generating human-like responses to text-based inputs. Its advanced natural language processing abilities have made it a valuable tool for a variety of applications, from customer service chatbots to language translation services. In general, similar to many other language models, ChatGPT uses an autoregression process to generate text. After training on massive data, ChatGPT predicts the likelihood of the next word in a sentence based on the previous words in the sentence. However, this autoregression working strategy of ChatGPT has caused some linguistic limitations. The first limitation is its tendency to generate text that exhibits a machine-patterned language. The second limitation is the generation of wrong invalid references. This paper conducts an experimental analysis of the linguistic limitations of ChatGPT caused by the word-by-word autoregression working strategy of ChatGPT. Three experiments will be conducted in this paper for proper evaluation: 1) Detecting texts written by ChatGPT. 2) Detecting text paraphrased by ChatGPT. 3) Detecting code written by ChatGPT. Extensive experiments involving a total number of 89,902 words and 552,381 characters have yielded compelling evidence that ChatGPT-generated text can be detected with a high percentage using AI techniques. Moreover, it was observed that AI-based detection methods did not consistently identify text paraphrased by ChatGPT due to the preservation of the original human structure within the paraphrased text. This also proves that there is a detectable difference between the human-based text structure and ChatGPT-based text structure.

It has recently been shown that there are no multimedia fingerprinting codes that can find all malicious users when they use arbitrary linear attacks plus adversarial noise. It is shown that such codes exist if the complete recovery property is limited to the IPP property, i.e., the property to find at least one malicious user. Moreover, we extend this property to a property that allows us to detect all users whose contribution to the forgery is large enough. Efficient decoding (tracing traitors) algorithms are developed for these codes.

This research constitutes a building block in a larger decision support system, which combines computer vision and multi-criteria decision making, to identify indigenous forage plants in close-range aerial images. Such a system enables managing the open pastures project of the Emirate of Sharjah, United Arab Emirates (UAE). The system aims at the identification of forage plants through the detection of plant inflorescences in depth images, and applying TOPSIS, to select and localize the best plant inflorescence.

Utilizing data redundancy has been widely proposed as a practical solution to hide a substantial amount of secret data. For digital images, the distribution of pixels in a 2D matrix has increased the possibility of having redundancy since each pixel has a relation with the eight neighboring pixels. However, the 3D nature of video segments has introduced an additional source of data redundancy, namely temporal redundancy. Several recent video steganography schemes have suggested exploiting temporal redundancy to hide the secret data. This paper aims to fully capture the redundancy available in a video segment by incorporating both spatial and temporal redundancy using an Octree segmentation method. The input video signal is segmented into homogeneous variable-sized 3D cubes to maximize the redundancy in each region and thus, stronger energy compaction is expected in the 3D Discrete Cosine Transform (3D-DCT) domain. In other words, the hiding capacity is maximized since most of the signal will be concentrated within a few significant 3D-DCT coefficients leaving a large area with insignificant 3DDCT coefficients, which can be safely replaced by the secret data without noticeable degradations in the quality of the stego signal.

This paper introduces velocity tracking control for differential wheeled mobile robot using vision information provided by a depth camera. This tracking control strategy is done into three phases. First, a depth camera mounted on the ceiling is used to collect texture and 3D shape information of the scene. Second, the piecewise 3D models generated in the previous phase and the Probabilistic Roadmap path planning algorithm (PRM) are used as an input for a robotic path planner to develop a free-obstacle path. Third, a kinematic controller is developed to track the desired path given from the path planner. Lyapunov theory is used to prove the stability of the closed loop system. Experimental results on a Pioneer P3-DX differential wheeled mobile robot and an Xtion PRO depth camera are illustrated at the end of the paper to prove the effectiveness and efficiency of the proposed tracking system.

Massively deployed inside Sony PS3 platforms, the STI Cell Broadband Engine is a multi-core processor with a PowerPC host processor (PPE) and 8 synergic processor engines (SPEs). In this paper, we describe three image processing applications which we implemented on the Cell BE. We report the performance measured on one Cell blade with varying numbers of synergic processor engines enabled, and with varying application parameters. These results were compared to the results obtained on the Cell BE running a single PPE and with all 8 SPEs disabled. The results reveal the superior performance of image processing and SIMD-prone applications on the 9-core Cell BE processor over conventional RISC processors.

Although existing robotic systems are interesting to interact with in the short term, it has been shown that after some weeks of quasi-regular encounters, humans gradually lose their interest, and meaningful longer-term human-robot relationships are not established. An underlying hypothesis driving the proposed project is that such relationships can be significantly enhanced if the human and the robot are gradually creating a pool of shared episodic memories that they can co-refer to, and if they are both embedded in a social web of other humans and robots they both know and encounter frequently. Thus, here we propose to use Facebook, a highly successful online networking resource for humans, towards enhancing longer-term human-robot relationships, by helping to address the above two prerequisites. As a starting point, we utilize social information in order to personalize human-robot dialogues, and to include references to past encounters and to encounters with friends within dialogues. A robot equipped with a modular software architecture (with IPC-intercommunicating modules for face recognition, a simple dialog system, a navigation subsystem, and a real-time Facebook connection/local social database) has been deployed, and is encountering humans in the environment of our lab. An early demonstration of a basic form of such encounters is shown in the submitted video. The system is expected to achieve two significant novelties: arguably being one of the first robots to be embedded in a social web, and being the first robot that can purposefully exploit and create social information that is available online. Furthermore, it is expected to provide empirical support for our main driving hypothesis, that the formation of shared episodic memories within a social web can lead to more meaningful long-term human-robot relationships.

Our project aims at supporting the creation of sustainable and meaningful longer-term human-robot relationships through the creation of embodied robots with face recognition and natural language dialogue capabilities, which exploit and publish social information available on the web (Facebook). Our main underlying experimental hypothesis is that such relationships can be significantly enhanced if the human and the robot are gradually creating a pool of shared episodic memories that they can co-refer to (shared memories), and if they are both embedded in a social web of other humans and robots they both know and encounter (shared friends). In this paper, we are presenting such a robot, which as we will see achieves two significant novelties.

A newly reported robust estimation approach for gray-scale image filtering is extended to address the more complex problem of denoising color television signals that have been degraded by weak-signal noise. This new technique differs from previous filtering methods in that it is geared towards the type of color degradations typical of weak-signal television reception, and thus takes into account the physical limitations associated with color signal transmission and the specific types of weak-signal noise generated. Experimental results demonstrating the effectiveness of this new technique are presented at the end of the paper.

This work develops a new robust statistical framework for blind image denoising. Robust statistics addresses the problem of estimation when the idealized assumptions about a system are occasionally violated. The contaminating noise in an image is considered as a violation of the assumption of spatial coherence of the image intensities and is treated as an outlier random variable. A denoised image is estimated by fitting a spatially coherent stationary image model to the available noisy data using a robust estimator-based regression method within an optimal-size adaptive window. The robust formulation aims at eliminating the noise outliers while preserving the edge structures in the restored image. Several examples demonstrating the effectiveness of this robust denoising technique are reported and a comparison with other standard denoising filters is presented.

We develop a vision system for highly mobile autonomous agents that is capable of dynamic obstacle avoidance. We demonstrate the robust performance of the system in artificial animals with directable, foveated eyes, situated in physics-based virtual worlds. Through active perception, each agent controls its eyes and body by continuously analyzing photorealistic binocular retinal image streams. The vision system computes stereo disparity and segments looming targets in the low-resolution visual periphery while controlling eye movements to track an object fixated in the high-resolution fovea. It matches segmented targets against mental models of colored objects of interest in order to decide whether the segmented objects are harmless or represent dangerous obstacles. The latter are localized enabling the artificial animal to exercise the sensorimotor control necessary to avoid collision.

We propose and demonstrate a new paradigm for active vision research that draws upon recent advances in the fields of artificial life and computer graphics. A software alternative to the prevailing hardware vision mindset, animat vision prescribes artificial animals, or animats, situated in physics-based virtual worlds as autonomous virtual robots possessing active perception systems. To be operative in its world, an animat must autonomously control its eyes and muscle-actuated body, applying computer vision algorithms to continuously analyze the retinal image streams acquired by its eyes in order to locomote purposefully through its world. We describe an initial animat vision implementation within lifelike artificial fishes inhabiting a physics-based, virtual marine world. Emulating the appearance, motion, and behavior of real fishes in their natural habitats, these animats are capable of spatially nonuniform retinal imaging, foveation, retinal image stabilization, color object recognition, and perceptually-guided navigation. These capabilities allow them to pursue moving targets such as fellow artificial fishes. Animat vision offers a fertile approach to the development, implementation, and evaluation of computational theories that profess sensorimotor competence for animal or robotic situated agents.>