Motion transfer has emerged as a promising direction for controllable video generation, yet existing methods largely focus on single-object scenarios and struggle when multiple objects require distinct motion patterns. In this work, we present FlexiMMT, the first implicit image-to-video (I2V) motion transfer framework that explicitly enables multi-object, multi-motion transfer. Given a static multi-object image and multiple reference videos, FlexiMMT independently extracts motion representations and accurately assigns them to different objects, supporting flexible recombination and arbitrary motion-to-object mappings. To address the core challenge of cross-object motion entanglement, we introduce a Motion Decoupled Mask Attention Mechanism that uses object-specific masks to constrain attention, ensuring that motion and text tokens only influence their designated regions. We further propose a Differentiated Mask Propagation Mechanism that derives object-specific masks directly from diffusion attention and progressively propagates them across frames efficiently. Extensive experiments demonstrate that FlexiMMT achieves precise, compositional, and state-of-the-art performance in I2V-based multi-object multi-motion transfer. Our project page is: https://ethan-li123.github.io/FlexiMMT_page/
The nuclear quadrupole collective states at low excitation energies are described in a novel, fully quantum mechanical and systematic manner as compared to traditional pictures initiated by Aage Bohr. The ellipsoidal shapes are shown to be triaxial in virtually all strongly deformed nuclei, in contrast to the Ansatz of axially symmetric shapes. The rotational bands of such triaxially deformed nuclei are described in a fully quantum mechanical way, i. e., without resorting to quantized free rotation of rigid body. The excitation energies within a rotational band, exhibiting the J (J + 1) dependence on angular momentum J, are shown to basically represent the change of binding energies due to nuclear forces. This differs from the interpretation á la Aage Bohr as rotational kinetic energies. The K quantum numbers are shown to be practically conserved for triaxial ellipsoids, which turned out to be a real but positive surprise to many people in the field. The so-called γ bands are shown to be K=2+ rotations rather than γ-vibrations, leading to a nice description of the so-called γγ 4+ state as a K=4+ rotation. Vibrational modes are also shown to emerge in this study. Thus, the whole picture of low-energy quadrupole collective motion of heavy nuclei has been renewed in a fully quantum mechanical fashion, which differs from the traditional picture but appears to be simpler and more natural.
Autonomous driving has seen significant progress, driven by extensive real-world data. However, in long-tail scenarios, accurately predicting the safety of the ego vehicle's future motion remains a major challenge due to uncertainties in dynamic environments and limitations in data coverage. In this work, we aim to explore whether it is possible to enhance the motion risk prediction capabilities of Vision-Language Models (VLM) by synthesizing high-risk motion data. Specifically, we introduce a Bird's-Eye View (BEV) based motion simulation method to model risks from three aspects: the ego-vehicle, other vehicles, and the environment. This allows us to synthesize plug-and-play, high-risk motion data suitable for VLM training, which we call DriveMRP-10K. Furthermore, we design a VLM-agnostic motion risk estimation framework, named DriveMRP-Agent. This framework incorporates a novel information injection strategy for global context, ego-vehicle perspective, and trajectory projection, enabling VLMs to effectively reason about the spatial relationships between motion waypoints and the environment. Extensive experiments demonstrate that by fine-tuning with DriveMRP-10K, our DriveMRP-Agent framework can significantly improve the motion risk prediction performance of multiple VLM baselines, with the accident recognition accuracy soaring from 27.13% to 88.03%. Moreover, when tested via zero-shot evaluation on an in-house real-world high-risk motion dataset, DriveMRP-Agent achieves a significant performance leap, boosting the accuracy from base_model's 29.42% to 68.50%, which showcases the strong generalization capabilities of our method in real-world scenarios.
This paper proposes an image-based robot motion planning method using a one-step diffusion model. While the diffusion model allows for high-quality motion generation, its computational cost is too expensive to control a robot in real time. To achieve high quality and efficiency simultaneously, our one-step diffusion model takes an approximately generated motion, which is predicted directly from input images. This approximate motion is optimized by additive noise provided by our novel noise optimizer. Unlike general isotropic noise, our noise optimizer adjusts noise anisotropically depending on the uncertainty of each motion element. Our experimental results demonstrate that our method outperforms state-of-the-art methods while maintaining its efficiency by one-step diffusion.
This article discusses how postfeminism plays into the exchange of romantic relationship advice between young women on the Chinese social media platform Douban. Employing Fairclough’s approach to critical discourse analysis, we elucidate how the process of giving and receiving romantic relationship advice facilitates the construction of a normative, postfeminist selfhood specific to China’s post-socialist transition. Under postfeminist aspirations for independence, self-empowerment, and personal success, this selfhood is constructed in social-mediated communication to offer an instrumentalist assessment of romantic relationships. In doing so, young Chinese women repudiate each other’s lovesickness, pathologizing romantic desires to achieve a false sense of emotional stability and mental strength. This gives rise to a girlfriend gaze on women’s emotional changes, establishing a self-/peer surveillance mechanism that regulates how members of the Douban community engage each other on the platform. While masked with a veneer of female empowerment and sisterly help, this mode of self-/peer surveillance not only perpetuates the stereotypical association of women with emotionality and irrationality but also marginalizes public discussions on structural gender inequalities, undermining collective efforts to challenge the patriarchal status quo. The research findings shed light on the functioning of neoliberal governance in contemporary China in the context of social-mediated communication.
This article probes the ways in which returning to Palestine is imagined in Razan AlSalah’s two video works Your Father Was Born 100 Years Old, and So Was the Nakba (2018) and Canada Park (2020). In foregrounding the refusal of configurations substantiated by state concessions and normalisation treaties, the article treats loss as central to the manifold rehearsals of return. In AlSalah’s work, loss is understood not as becoming less, but rather as a proposition for becoming otherwise. Here, the practice of loss is explored through the glitch as both a conceptual framework espousing opacity and a pragmatic tool engaging pixel breaks. Rather than reducing the glitch to a mere erroneous aesthetic, the article underscores the active capacity in encountering a glitch or deliberately engendering it by exploring the tensions between colonial imageries reproduced in digital maps, and, in contrast, montage as AlSalah’s tool for intervention. Finally, the article serves as a theoretical experimentation with what I call “dialectical poethics,” which reads the filmic return through loss as an attempt to go against linearity, intelligibility, and finality, yet insists on a materialist grounding of the glitch as a method that is historically situated rather than always-already emancipatory.
Conventional Task and Motion Planning (TAMP) approaches rely on manually crafted interfaces connecting symbolic task planning with continuous motion generation. These domain-specific and labor-intensive modules are limited in addressing emerging tasks in real-world settings. Here, we present LLM^3, a novel Large Language Model (LLM)-based TAMP framework featuring a domain-independent interface. Specifically, we leverage the powerful reasoning and planning capabilities of pre-trained LLMs to propose symbolic action sequences and select continuous action parameters for motion planning. Crucially, LLM^3 incorporates motion planning feedback through prompting, allowing the LLM to iteratively refine its proposals by reasoning about motion failure. Consequently, LLM^3 interfaces between task planning and motion planning, alleviating the intricate design process of handling domain-specific messages between them. Through a series of simulations in a box-packing domain, we quantitatively demonstrate the effectiveness of LLM^3 in solving TAMP problems and the efficiency in selecting action parameters. Ablation studies underscore the significant contribution of motion failure reasoning to the success of LLM^3. Furthermore, we conduct qualitative experiments on a physical manipulator, demonstrating the practical applicability of our approach in real-world settings.
Scholars debate and categorize varieties of bigotry in terms of responsibility, whether individual, institutional, structural, systematic, or systemic. Using the distance of fantasy narrative and “monsterism” in specific cases in Buffy the Vampire Slayer, the authors argue for the need to consider both individual and systemic bigotry, avoiding reductionist views and thus making it possible to more successfully fight bigotry.
We construct an analogue of Dyson Brownian motion in the Siegel half-space H that we term Siegel Brownian motion. Given βin (0,\infty], a stochastic flow for Z_t in H is introduced so that the law of the eigenvalues λ_t of the cross ratio matrix R(Z_t,iI_n) is determined by the Ito differential equation corresponds to stochastic gradient ascent of a function S. S turns out to be the log volume of isospectral orbit in H and can be understood as a Boltzmann entropy. In the limit β=\infty, the group orbits evolve by motion by minus a half times mean curvature.
Since the invention of dynamic diode, its physical properties and potential applications have attracted wide attentions. A lot of attempts have been made to harvest the rebounding current and voltage of dynamic diode. However, the underlying physical mechanism of its carrier transport characteristic was rarely explored carefully. Here, the electrical transport properties of the dynamic diode are systematically investigated with a mechanical motion tuned method, where the dynamic current-voltage curve shows a gentler growth trend compared to the static curve. The rebounding current increases with motion velocity and contact force, resulting in a reduced current with the same bias voltage and an oscillation current with a changing velocity and force. Besides, we propose a circuit model with an accurate mathematical formula expression to describe the oscillation current, where an imaginary parameter n0 is creatively added to the exponential growth term. This work shows a physical picture of adjust microscopic carrier motion with macroscopic mechanical motion, which provides strong theoretical support for designing dynamic diode devices with better performance in the future.
The Arctic panorama has often been framed by conversations of the Victorian imperial imaginary and was originally conceived to showcase anglophone exploration, heroism, and scientific discovery. This paper explores the complementary and multifaceted visual representations of Arctic glacial ice, from the Victorian spectacle of the panorama to contemporary virtual reality technology. I look at how glaciers have been depicted, documented, and presented within panoramic media over the past two centuries. I explore how ice moves through both time and space, confronting climate histories within physical and spatiotemporal ideas of movement. Alongside the materiality of ice, I consider the modes of observation involved in creating and viewing these panoramic pictures. In making Arctic exploratory history tangible and visual, I do not seek to simply revive a sublime imaginary and document ice loss but look to the panorama as a method of engaging polar exploration and scientific discovery in the study of visual culture.
Directed by Maggie Gyllenhaal, the movie The lost daughter, adapted from homonymous novel written by Elena Ferrante, narrates Leda´s vacation, a College Professor who questions her own maternity when she gets in touch with a young mother and her daughter. Inside an uncomfortable game of mirrors, the protagonist dives into her past, gradually exposing her biography, surprising the spectator with family revelations. This review proposal is to analyse the film in the key of otherness and discuss the characters motivations from the meeting and the generations conflict, especially regarding to motherhood ambivalences perspective. Topics such as aging, female sexual freedom, marriage, emerge from the cinematographic narrative, released in 2021, and starring Olivia Colman. One of the goals is to examine the film using the book The second sex, by Simone de Beauvoir (2019), for whom a woman is not born a woman, she becomes a woman, since the social pressures and the adjustments demanded by the patriarchal configuration delineate the feminine.
Nosso percurso elegeu filmes capazes de pontuar a discussão sobre a dificuldade material e a resiliência nas situações-limite impostas a cidadãos em espaços sensivelmente modificados por novas condições de trabalho e sobrevivência. Partimos das escolhas narrativas do cinema italiano, destacando filmes de Gianni Amelio e sua problematização das migrações para alcançar o sentido experimental no documentário de Gianfranco Rosi, hábil em revisar a periferia romana trazendo a vida excluída e sobrevivente para o primeiro plano. Chloé Zhao aproveita o livro de Jessica Bruder e realiza o filme que atualiza as dimensões antropológica e audiovisual por meio de indivíduos praticamente apagados da existência civil. Como "Sacro GRA", "Nomadland" conta histórias de pessoas forçadas a ocupar territórios inabitáveis em seu próprio país. Ressaltando sequências originais por meio da análise imanente, investigamos como os conceitos de território e espaço são revirados pela própria composição dos filmes. As abordagens pedem novos modos de filmar e solicitam do espectador uma revisão urgente de conceitos políticos, econômicos e sociais, tanto quanto uma renovação do olhar para os filmes, para visualizar com profundidade os espaços ocupados pelos invisíveis, os limites e os modos de extravasar as condições de trabalho nas histórias contadas e na própria filmagem.
This article reflects on the sixty years marking the anniversary of the encyclical letter Miranda Prorsus by Pope Pius XII. Miranda Prorsus was the first document written in 1957 by the Church to refl ect on the three important means of communication: Motion Pictures (Film), Television (TV) and Radio. It highlighted the importance of these “remarkable technical inventions” to aid humanity in as far as development and understanding the media was concerned. Each of these three instruments of communication is examined in both the strengths and weaknesses they carry, but much more, how they can play a role in advancing humanity in the area of morality and truth telling. The article tries to use some of the important highlights in the context of Zambia my country that has embraced these means of communication with radio stations set up by the respective Bishops and a Television soon to be launched by the Zambia Conference of Catholic Bishops (ZCCB). In retrospect the article tries to show how these means and technical inventions can become handy in as far as the evangelization is concerned.
Ethnology. Social and cultural anthropology, Practical Theology
The analysis of human motion opens up a wide range of possibilities, such as realistic training simulations or authentic motions in robotics or animation. One of the problems underlying motion analysis is the meaningful comparison of actions based on similarity measures. Since the motion analysis is application-dependent, it is essential to find the appropriate motion similarity method for the particular use case. This state of the art report provides an overview of human motion analysis and different similarity modeling methods, while mainly focusing on approaches that work with 3D motion data. The survey summarizes various similarity aspects and features of motion and describes approaches to measuring the similarity between two actions.
Igor Slinko, Anna Vorontsova, Filipp Konokhov
et al.
Optical Flow (OF) and depth are commonly used for visual odometry since they provide sufficient information about camera ego-motion in a rigid scene. We reformulate the problem of ego-motion estimation as a problem of motion estimation of a 3D-scene with respect to a static camera. The entire scene motion can be represented as a combination of motions of its visible points. Using OF and depth we estimate a motion of each point in terms of 6DoF and represent results in the form of motion maps, each one addressing single degree of freedom. In this work we provide motion maps as inputs to a deep neural network that predicts 6DoF of scene motion. Through our evaluation on outdoor and indoor datasets we show that utilizing motion maps leads to accuracy improvement in comparison with naive stacking of depth and OF. Another contribution of our work is a novel network architecture that efficiently exploits motion maps and outperforms learnable RGB/RGB-D baselines.
In this paper, we are interested in self-supervised learning the motion cues in videos using dynamic motion filters for a better motion representation to finally boost human action recognition in particular. Thus far, the vision community has focused on spatio-temporal approaches using standard filters, rather we here propose dynamic filters that adaptively learn the video-specific internal motion representation by predicting the short-term future frames. We name this new motion representation, as dynamic motion representation (DMR) and is embedded inside of 3D convolutional network as a new layer, which captures the visual appearance and motion dynamics throughout entire video clip via end-to-end network learning. Simultaneously, we utilize these motion representation to enrich video classification. We have designed the frame prediction task as an auxiliary task to empower the classification problem. With these overall objectives, to this end, we introduce a novel unified spatio-temporal 3D-CNN architecture (DynamoNet) that jointly optimizes the video classification and learning motion representation by predicting future frames as a multi-task learning problem. We conduct experiments on challenging human action datasets: Kinetics 400, UCF101, HMDB51. The experiments using the proposed DynamoNet show promising results on all the datasets.
The prolepsis in Denis Villeneuve's Arrival emphasises the cyclical nature of the film's narrative and anchors human reproduction as a central theme. Pregnancy, the pregnant body, and the physical, experiential nature of birth, commonly heavily gendered in film, are misleading focal points in the narrative. The presence of the unborn as a subtext in the film problematises Iris Marion Young's (2005) notion of pregnant embodiment as a subjective lived-body experience. The viewer is encouraged to empathise with the complexity of birth, life, and death as part of Louise's lived-body experience, but is finally confronted with the uncertainty of maternity, pregnancy and the unborn. When Barbara Duden (1992) calls the unborn foetus a “not-yet”, she describes the process by which the foetus achieves a legal status, and the precarious nature of ascribing life or personhood. The prolepsis, which punctuates the main narrative, emphasises the reversibility and irreversibility of life that does “not-yet” exist. Importantly, the constant hovering over the threshold of life in the film complicates the timeline of reproduction. At the end of the film's narrative, the main character Louise Banks (Amy Adams), is “not-yet” pregnant, is “not-yet” a parent, and has “not-yet” lost a child. The temporal shifts in the film rely on repositioning or reorienting both Louise and the viewer to the “not-yet” reproductive body and the “not-yet” child. By presenting events out of chronological time and returning to the time before and after a child is born, the film ultimately raises crucial questions about the ethics of reproduction, the quality of life, and issues of consent.