Hasil untuk "Motion pictures"

Yiyang Cao, Yunze Deng, Ziyu Lin et al.

Text-to-motion generation, a rapidly evolving field in computer vision, aims to produce realistic and text-aligned motion sequences. Current methods primarily focus on spatial-temporal modeling or independent frequency domain analysis, lacking a unified framework for joint optimization across spatial, temporal, and frequency domains. This limitation hinders the model's ability to leverage information from all domains simultaneously, leading to suboptimal generation quality. Additionally, in motion generation frameworks, motion-irrelevant cues caused by noise are often entangled with features that contribute positively to generation, thereby leading to motion distortion. To address these issues, we propose Tri-Domain Causal Text-to-Motion Generation (TriC-Motion), a novel diffusion-based framework integrating spatial-temporal-frequency-domain modeling with causal intervention. TriC-Motion includes three core modeling modules for domain-specific modeling, namely Temporal Motion Encoding, Spatial Topology Modeling, and Hybrid Frequency Analysis. After comprehensive modeling, a Score-guided Tri-domain Fusion module integrates valuable information from the triple domains, simultaneously ensuring temporal consistency, spatial topology, motion trends, and dynamics. Moreover, the Causality-based Counterfactual Motion Disentangler is meticulously designed to expose motion-irrelevant cues to eliminate noise, disentangling the real modeling contributions of each domain for superior generation. Extensive experimental results validate that TriC-Motion achieves superior performance compared to state-of-the-art methods, attaining an outstanding R@1 of 0.612 on the HumanML3D dataset. These results demonstrate its capability to generate high-fidelity, coherent, diverse, and text-aligned motion sequences. Code is available at: https://caoyiyang1105.github.io/TriC-Motion/.

Elisa Tosello, Arthur Bit-Monnot, Davide Lusuardi et al.

Task and Motion Planning combines high-level task sequencing (what to do) with low-level motion planning (how to do it) to generate feasible, collision-free execution plans. However, in many real-world domains, such as automated warehouses, tasks are predefined, shifting the challenge to if, when, and how to execute them safely and efficiently under resource, time and motion constraints. In this paper, we formalize this as the Scheduling and Motion Planning problem for multi-object navigation in shared workspaces. We propose a novel solution framework that interleaves off-the-shelf schedulers and motion planners in an incremental learning loop. The scheduler generates candidate plans, while the motion planner checks feasibility and returns symbolic feedback, i.e., spatial conflicts and timing adjustments, to guide the scheduler towards motion-feasible solutions. We validate our proposal on logistics and job-shop scheduling benchmarks augmented with motion tasks, using state-of-the-art schedulers and sampling-based motion planners. Our results show the effectiveness of our framework in generating valid plans under complex temporal and spatial constraints, where synchronized motion is critical.

Francesco Sticchi

Haipeng Li, Tianhao Zhou, Zhanglei Yang et al.

Estimating 2D camera motion is a fundamental computer vision task that models the projection of 3D camera movements onto the 2D image plane. Current methods rely on either homography-based approaches, limited to planar scenes, or meshflow techniques that use grid-based local homographies but struggle with complex non-linear transformations. A key insight of our work is that combining flow fields from different homographies creates motion patterns that cannot be represented by any single homography. We introduce CamFlow, a novel framework that represents camera motion using hybrid motion bases: physical bases derived from camera geometry and stochastic bases for complex scenarios. Our approach includes a hybrid probabilistic loss function based on the Laplace distribution that enhances training robustness. For evaluation, we create a new benchmark by masking dynamic objects in existing optical flow datasets to isolate pure camera motion. Experiments show CamFlow outperforms state-of-the-art methods across diverse scenarios, demonstrating superior robustness and generalization in zero-shot settings. Code and datasets are available at our project page: https://lhaippp.github.io/CamFlow/.

Tri Tung Nguyen Nguyen, Quang Tien Dam, Dinh Tuan Tran et al.

Effective human behavior modeling is critical for successful human-robot interaction. Current state-of-the-art approaches for predicting listening head behavior during dyadic conversations employ continuous-to-discrete representations, where continuous facial motion sequence is converted into discrete latent tokens. However, non-verbal facial motion presents unique challenges owing to its temporal variance and multi-modal nature. State-of-the-art discrete motion token representation struggles to capture underlying non-verbal facial patterns making training the listening head inefficient with low-fidelity generated motion. This study proposes a novel method for representing and predicting non-verbal facial motion by encoding long sequences into a sparse sequence of keyframes and transition frames. By identifying crucial motion steps and interpolating intermediate frames, our method preserves the temporal structure of motion while enhancing instance-wise diversity during the learning process. Additionally, we apply this novel sparse representation to the task of listening head prediction, demonstrating its contribution to improving the explanation of facial motion patterns.

Nicola Bamonti

Inertia has long been treated as the paradigm of natural motion. This paper challenges this identification through the lens of General Relativity. Drawing on Norton (2012)'s distinction between idealisation and approximation and analysing key results from Tamir (2012) on the theorems of Geroch-Jang, Ehlers-Geroch, Einstein-Grommer, and Geroch-Traschen, I argue that geodesic motion -- commonly treated as the relativistic expression of inertia -- fails to qualify as either. Rather, geodesic motion is best understood as a useful construct -- a formal artefact of the theory's geometric structure, without real or fictitious instantiation, and excluded by the dynamical structure of General Relativity. In place of inertial motion, I develop a layered account of natural motion, which is not encoded in a single "master equation of motion." Extended, structured, and backreacting bodies require successively refined dynamical formalisms that systematically depart from geodesic motion. This pluralist framework displaces geodesic motion as the privileged expression of pure gravitational motion, replacing it with a dynamically grounded hierarchy of approximations fully consistent with the Einstein field equations. Inertial motion thus emerges not as the universal default of motion under gravity alone, but as a formal construct that stands apart from the pluralistic framework in which natural motion is genuinely realised.

Marina Ritthaler, Andy Regensky, André Kaup

Efficient compression of 360-degree video content requires the application of advanced motion models for interframe prediction. The Motion Plane Adaptive (MPA) motion model projects the frames on multiple perspective planes in the 3D space. It improves the motion compensation by estimating the motion on those planes with a translational diamond search. In this work, we enhance this motion model with an affine parameterization and motion estimation method. Thereby, we find a feasible trade-off between the quality of the reconstructed frames and the computational cost. The affine motion estimation is hereby done with the inverse compositional Lucas-Kanade algorithm. With the proposed method, it is possible to improve the motion compensation significantly, so that the motion compensated frame has a Weighted-to-Spherically-uniform Peak Signal-to-Noise Ratio (WS-PSNR) which is about 1.6 dB higher than with the conventional MPA. In a basic video codec, the improved inter prediction can lead to Bjøntegaard Delta (BD) rate savings between 9 % and 35 % depending on the block size (BS) and number of motion parameters.

Rhea Malhotra, William Chong, Catie Cuan et al.

Generating sequences of human-like motions for humanoid robots presents challenges in collecting and analyzing reference human motions, synthesizing new motions based on these reference motions, and mapping the generated motion onto humanoid robots. To address these issues, we introduce SynSculptor, a humanoid motion analysis and editing framework that leverages postural synergies for training-free human-like motion scripting. To analyze human motion, we collect 3+ hours of motion capture data across 20 individuals where a real-time operational space controller mimics human motion on a simulated humanoid robot. The major postural synergies are extracted using principal component analysis (PCA) for velocity trajectories segmented by changes in robot momentum, constructing a style-conditioned synergy library for free-space motion generation. To evaluate generated motions using the synergy library, the foot-sliding ratio and proposed metrics for motion smoothness involving total momentum and kinetic energy deviations are computed for each generated motion, and compared with reference motions. Finally, we leverage the synergies with a motion-language transformer, where the humanoid, during execution of motion tasks with its end-effectors, adapts its posture based on the chosen synergy. Supplementary material, code, and videos are available at https://rhea-mal.github.io/humanoidsynergies.io.

Pierre-Louis Phan, Nicolas Rambaux

Five Venus missions are under development to study the planet in the next decade, with both NASA's VERITAS and ESA's EnVision featuring a geophysical investigation among their objectives. Their radar and gravity experiments will determine Venus's orientation, enabling analyses of its spin dynamics to infer relevant geophysical and atmospheric properties. This work aims to characterize Venus's polar motion, defined as the motion of its spin axis in a body-fixed frame. We focus on signatures from its interior and atmosphere to support potential detections of polar motion by future orbiters. We developed a polar motion model for a triaxial planet accounting for solar torque, centrifugal and tidal deformations of a viscoelastic mantle, and atmospheric dynamics. Core-mantle coupling effects were analyzed separately, considering a simplified spherical core. We computed the period and damping time of the free motion (i.e., the Chandler wobble) and determined the frequencies and amplitudes of the forced motion. We revisited the Chandler frequency expression. Solar torque is the dominant phenomenon affecting Venus's Chandler frequency, increasing it by a factor of 2.75. Our model predicts a Chandler period in the range [12900 ; 18900] years. The Chandler wobble appears as a linear polar drift of about 90 meters on Venus's surface during EnVision's 4-year primary mission, at the limit of its resolution. We also predict forced polar motion oscillations with an amplitude of about 20 meters, driven by the atmosphere and the solar torque. Compared to the 240 meter spin axis precession occurring in inertial space over this duration, these results suggest that Venus's polar motion could also be detectable by future orbiters. Polar motion should be incorporated into rotation models when anticipating these missions, providing additional constraints on the interior structure of Venus.

Sofia Afonso Lopes, Ana Paula Tavares

Estreado em 2004, num momento de viragem do cinema angolano, o filme Na Cidade Vazia (2004) de Maria João Ganga narra a estória de N’dala, uma criança de onze anos que, em 1991, aterra em Luanda em fuga da guerra civil que vitimara, no Bié, a sua família. Na capital, N’dala trava amizade com Zé, um jovem que protagoniza, numa peça escolar, Ngunga, o herói da célebre obra de estreia do escritor Pepetela. Num processo de deambulação pelas ruas luandenses, N’dala personifica a dor da perda e da deslocação forçada causadas pelo conflito intestino, evocando, simultaneamente e pelas suas semelhanças com Ngunga, os efeitos dilatados e na longue-durée do colonialismo e da luta de libertação pela independência do território. Os supracitados eventos históricos, plasmados alegoricamente na tela daquela que foi a primeira longa-metragem assinada por uma mulher angolana, são convocados enquanto objecto de análise para pensar a experiência do luto, mas também o sentido comunal que do trauma da(s) guerra(s) pode germinar. Neste quadro, a experiência feminina – activada e actualizada na tela pelas mulheres que participam na narrativa – assume particular destaque.

Céu e Silva, João

Director José de Pina premiered the film The Mystery of Boca do Inferno in 1989, which explores the tempestuous relationship between poet Fernando Pessoa and magician Aleister Crowley. Their encounter took place in 1930 and ended with Crowley's fake suicide, sparking numerous newspaper articles and police investigations in both England and Portugal. The film revisits this episode, drawing on Pessoa's penchant for detective stories and both men's interest in mediumistic faculties.

Luozhou Wang, Ziyang Mai, Guibao Shen et al.

In this work, we present a novel approach for motion customization in video generation, addressing the widespread gap in the exploration of motion representation within video generative models. Recognizing the unique challenges posed by the spatiotemporal nature of video, our method introduces Motion Embeddings, a set of explicit, temporally coherent embeddings derived from a given video. These embeddings are designed to integrate seamlessly with the temporal transformer modules of video diffusion models, modulating self-attention computations across frames without compromising spatial integrity. Our approach provides a compact and efficient solution to motion representation, utilizing two types of embeddings: a Motion Query-Key Embedding to modulate the temporal attention map and a Motion Value Embedding to modulate the attention values. Additionally, we introduce an inference strategy that excludes spatial dimensions from the Motion Query-Key Embedding and applies a differential operation to the Motion Value Embedding, both designed to debias appearance and ensure the embeddings focus solely on motion. Our contributions include the introduction of a tailored motion embedding for customization tasks and a demonstration of the practical advantages and effectiveness of our method through extensive experiments.

Lilly Joan Gutzeit, Victor Tiberius

The motion picture industry is subject to extensive business and management research conducted on a wide range of topics. Due to high research productivity, it is challenging to keep track of the abundance of publications. Against this background, we employ a bibliographic coupling analysis to gain a comprehensive understanding of current research topics. The following themes were defined: Key factors for success, word of mouth and social media, organizational and pedagogical dimensions, advertising—product placement and online marketing, tourism, the influence of data, the influence of culture, revenue maximization and purchase decisions, and the perception and identification of audiences. Based on the cluster analysis, we suggest the following future research opportunities: Exploring technological innovations, especially the influence of social media and streaming platforms in the film industry; the in-depth analysis of the use of artificial intelligence in film production, both in terms of its creative potential and ethical and legal challenges; the exploration of the representation of wokeness and minorities in films and their cultural and economic significance; and, finally, a detailed examination of the long-term effects of the COVID-19 pandemic and other crises on the film industry, especially in terms of changed consumption habits and structural adjustments.

Sigal Raab, Inbal Leibovitch, Guy Tevet et al.

Synthesizing realistic animations of humans, animals, and even imaginary creatures, has long been a goal for artists and computer graphics professionals. Compared to the imaging domain, which is rich with large available datasets, the number of data instances for the motion domain is limited, particularly for the animation of animals and exotic creatures (e.g., dragons), which have unique skeletons and motion patterns. In this work, we present a Single Motion Diffusion Model, dubbed SinMDM, a model designed to learn the internal motifs of a single motion sequence with arbitrary topology and synthesize motions of arbitrary length that are faithful to them. We harness the power of diffusion models and present a denoising network explicitly designed for the task of learning from a single input motion. SinMDM is designed to be a lightweight architecture, which avoids overfitting by using a shallow network with local attention layers that narrow the receptive field and encourage motion diversity. SinMDM can be applied in various contexts, including spatial and temporal in-betweening, motion expansion, style transfer, and crowd animation. Our results show that SinMDM outperforms existing methods both in quality and time-space efficiency. Moreover, while current approaches require additional training for different applications, our work facilitates these applications at inference time. Our code and trained models are available at https://sinmdm.github.io/SinMDM-page.

Bo Han, Hao Peng, Minjing Dong et al.

Human motion generation aims to produce plausible human motion sequences according to various conditional inputs, such as text or audio. Despite the feasibility of existing methods in generating motion based on short prompts and simple motion patterns, they encounter difficulties when dealing with long prompts or complex motions. The challenges are two-fold: 1) the scarcity of human motion-captured data for long prompts and complex motions. 2) the high diversity of human motions in the temporal domain and the substantial divergence of distributions from conditional modalities, leading to a many-to-many mapping problem when generating motion with complex and long texts. In this work, we address these gaps by 1) elaborating the first dataset pairing long textual descriptions and 3D complex motions (HumanLong3D), and 2) proposing an autoregressive motion diffusion model (AMD). Specifically, AMD integrates the text prompt at the current timestep with the text prompt and action sequences at the previous timestep as conditional information to predict the current action sequences in an iterative manner. Furthermore, we present its generalization for X-to-Motion with "No Modality Left Behind", enabling the generation of high-definition and high-fidelity human motions based on user-defined modality input.

Mathias S. Feinler, Bernadette N. Hahn

Magnetic Resonance Imaging allows high resolution data acquisition with the downside of motion sensitivity due to relatively long acquisition times. Even during the acquisition of a single 2D slice, motion can severely corrupt the image. Retrospective motion correction strategies do not interfere during acquisition time but operate on the motion affected data. Known methods suited to this scenario are compressed sensing (CS), generative adversarial networks (GANs), and motion estimation. In this paper we propose a strategy to correct for motion artifacts using Deep Convolutional Neuronal Networks (Deep CNNs) in a reliable and verifiable manner by explicit motion estimation. The sensitivity encoding (SENSE) redundancy that multiple receiver coils provide, has in the past been used for acceleration, noise reduction and rigid motion compensation. We show that using Deep CNNs the concepts of rigid motion compensation can be generalized to more complex motion fields. Using a simulated synthetic data set, our proposed supervised network is evaluated on motion corrupted MRIs of abdomen and head. We compare our results with rigid motion compensation and GANs.

Sihan Ma, Qiong Cao, Hongwei Yi et al.

Demystifying complex human-ground interactions is essential for accurate and realistic 3D human motion reconstruction from RGB videos, as it ensures consistency between the humans and the ground plane. Prior methods have modeled human-ground interactions either implicitly or in a sparse manner, often resulting in unrealistic and incorrect motions when faced with noise and uncertainty. In contrast, our approach explicitly represents these interactions in a dense and continuous manner. To this end, we propose a novel Ground-aware Motion Model for 3D Human Motion Reconstruction, named GraMMaR, which jointly learns the distribution of transitions in both pose and interaction between every joint and ground plane at each time step of a motion sequence. It is trained to explicitly promote consistency between the motion and distance change towards the ground. After training, we establish a joint optimization strategy that utilizes GraMMaR as a dual-prior, regularizing the optimization towards the space of plausible ground-aware motions. This leads to realistic and coherent motion reconstruction, irrespective of the assumed or learned ground plane. Through extensive evaluation on the AMASS and AIST++ datasets, our model demonstrates good generalization and discriminating abilities in challenging cases including complex and ambiguous human-ground interactions. The code will be available at https://github.com/xymsh/GraMMaR.

Gabriele

"Sulle tracce del crimine. Viaggio nel giallo e nero Rai / Hot on the Heels of Crime. A Journey into the World of Detectives on Rai" (Roma, Museo di Roma in Trastevere October 7, 2020 – March 31, 2021) Exhibition and Catalogue Roma: RaiLibri, 2020

Federico Pagello

Denys Rozumnyi, Martin R. Oswald, Vittorio Ferrari et al.

We propose a method for jointly estimating the 3D motion, 3D shape, and appearance of highly motion-blurred objects from a video. To this end, we model the blurred appearance of a fast moving object in a generative fashion by parametrizing its 3D position, rotation, velocity, acceleration, bounces, shape, and texture over the duration of a predefined time window spanning multiple frames. Using differentiable rendering, we are able to estimate all parameters by minimizing the pixel-wise reprojection error to the input video via backpropagating through a rendering pipeline that accounts for motion blur by averaging the graphics output over short time intervals. For that purpose, we also estimate the camera exposure gap time within the same optimization. To account for abrupt motion changes like bounces, we model the motion trajectory as a piece-wise polynomial, and we are able to estimate the specific time of the bounce at sub-frame accuracy. Experiments on established benchmark datasets demonstrate that our method outperforms previous methods for fast moving object deblurring and 3D reconstruction.