Blind or Low-Vision (BLV) users often rely on audio descriptions (AD) to access video content. However, conventional static ADs can leave out detailed information in videos, impose a high mental load, neglect the diverse needs and preferences of BLV users, lack immersion. To tackle these challenges, we introduce Spica, an AI-powered system that enables interactively explore Informed by prior empirical studies consumption, Spica offers interactive mechanisms for supporting temporal navigation...

Realistic 3D facial generation based on Neural Radiance Fields (NeRFs) from 2D sketches benefits various applications. Despite the high realism of free-view rendering results NeRFs, it is tedious and difficult for artists to achieve detailed control manipulation. Meanwhile, due its conciseness expressiveness, sketching has been widely used image editing. Applying NeRFs challenging inherent uncertainty with constraints, a significant gap in content richness when generating faces sparse...

Recent methods for synthesizing 3D-aware face images have achieved rapid development thanks to neural radiance fields, allowing high quality and fast inference speed. However, existing solutions editing facial geometry appearance independently usually require retraining are not optimized the recent work of generation, thus tending lag behind generation process. To address these issues, we introduce NeRFFaceEditing, which enables decoupling in pretrained tri-plane-based field while retaining...

3D-aware portrait lighting control is an emerging and promising domain, thanks to the recent advance of generative adversarial networks neural radiance fields. Existing solutions typically try decouple from geometry appearance for disentangled with explicit representation (e.g., Lambertian or Phong). However, they either are limited a constrained condition directional light) demand tricky-to-fetch dataset as supervision intrinsic compositions albedo). We propose NeRFFaceLighting explore...

The self-assembly of cuprous iodide supramolecular clusters with novel molecular structures and excellent photocatalytic performance through weak noncovalent interactions remains a major challenge. In this work, we fabricate dimeric anionic cluster [(Cu8I14)6–(L13+)2] (HUBU-1) using multicomponent confined strategy via host–guest encapsulation associated multicharge electrostatic interaction. Furthermore, HUBU-1 exhibits exceptional visible light absorption capability surmountable energy...

Spatial audio offers more immersive video consumption experiences to viewers; however, creating and editing spatial often expensive requires specialized hardware equipment skills, posing a high barrier for amateur creators. We present Mimosa, human-AI co-creation tool that enables users computationally generate manipulate effects. For with only monaural or stereo audio, Mimosa automatically grounds each sound source the corresponding sounding object in visual scene further validate fix...

This paper presents a novel time-optimal motion planning strategy for mobile robot with kinematic constraints. The method works in environments presence of obstacles, without needing to generate the configuration space robot. Further, it derives minimum time first derivative smooth path, as opposed distance path which is commonly given by various present solution techniques. problem solved three stages: (i) A reduced visibility graph point object obtained. (ii) converted into feasible...

To solve the problems that basic ant colony algorithm (ACO) is easy to fall into local optimum, path too long and there are many turns when searching of mobile robot, an improved ACO proposed in this paper. The randomly sets walkable position points obstacle-free area map, so ants reach next at any angle according proximity point. Then we incorporate idea A* valuation function heuristic let drive point on line between starting ending with a greater probability. simulation results indicate...

Finding a three dimensional shortest path is of importance in the development automatic planning for mobile robots and robot manipulators, practical implementation, algorithms need to be efficient. Presented method three-dimensional space presence convex polyhedra. It based on visibility graph approach, extended from two space. A collineation introduced identification visible edges graph. The principle minimum potential energy adopted finding set sub-shortest paths via different edge...

An automated parallel parking strategy for a car-like mobile robot is presented. The study considers general cases of rectangular within space. system works in three phases. In scanning phase the environment detected by ultrasonic sensors mounted on and position manoeuvring path produced if space sufficient. Then positioning reverses to edge space, avoiding potential collisions. Finally, moves unified pattern, which may require backward forward manoeuvres depending dimensions Motion...

This paper presents a new motion planning approach, including reversal manoeuvres, for car-like robots subject to non-holonomic constraints. The complete path algorithm and describes the procedure constructing collision-free mobile robot constrained by its rectangular shape kinematics. Emphasis is made on techniques of transfers, which are used compensate limited manoeuvrability robot. approach works entirely in workspace, as opposed building higher dimensional configuration space (C-space)....

Abmcf- Presented is a novel motion strategy for mobile, car like robot that subject to kinematic constraints. The algorithm operates on the original obstacles, without needing generate configuration space obstacles dimensioned robot. path generated in three stages: (i) shortest problem point solved; (ii) free relative locally evaluated by minimum distance computations; (iii) modified account size and constraints of mobile If fails be into feasible robot, process repeated with second...

Synthesizing realistic videos of talking faces under custom lighting conditions and viewing angles benefits various downstream applications like video conferencing. However, most existing relighting methods are either time-consuming or unable to adjust the viewpoints. In this paper, we present first real-time 3D-aware method for in-the-wild based on Neural Radiance Fields (NeRF). Given an input portrait video, our can synthesize both novel views with a photo-realistic disentangled 3D...

Soft robots, while highly adaptable to diverse environments through various actuation methods, still face significant performance boundary due the inherent properties of materials. These limitations manifest in challenge guaranteeing rapid response and large-scale movements simultaneously, ultimately restricting robots' absolute speed overall efficiency. In this paper, we introduce a high-frequency pneumatic oscillator (HIPO) overcome these challenges. Through collision-induced phase...

Geometric reconstruction of opaque surfaces from images is a longstanding challenge in computer vision, with renewed interest volumetric view synthesis algorithms using radiance fields. We leverage the geometry field proposed recent work for stochastic surfaces, which can then be converted to volume densities. adapt Gaussian kernels or surfels splat rather than volume, enabling precise solids. Our first contribution derive an efficient and almost exact differentiable rendering algorithm...

Bio-robot is a kind of creature controlled by human beings applying intervention signals through control technology to regulate biological behavior. The animal can be artificially and become robot stimulating the brain motor area animal. To realize precise carp robot, it necessary monitor its motion state quantify Therefore, article proposes method computer vision detect effect robot. In this study, combining inter-frame difference with Gaussian mixture background modeling was used extract...

Novel view synthesis from a sparse set of input images is challenging problem great practical interest, especially when camera poses are absent or inaccurate. Direct optimization and usage estimated depths in neural radiance field algorithms usually do not produce good results because the coupling between depths, inaccuracies monocular depth estimation. In this paper, we leverage recent 3D Gaussian splatting method to develop novel construct-and-optimize for without poses. Specifically,...