arXiv subjects

Amit H. Bermano

Publications and source records attributed to Amit H. Bermano.

At least 19 recordsLinked to original sources

arXiv2026

Generalized Audio-Driven Synthesis of Precise Drummer Motion

Music-driven character animation enables and enhances transformative applications in entertainment and interactive education. However, synthesizing realistic drumming motion from audio remains challenging due to the inherent tension between high-acceleration dynamics and the need for extreme spatial-temporal precision. Existing approaches, often reliant on motion matching or MIDI input, struggle with generalizing to diverse real-world audio. Moreover, the field lacks standardized evaluation metrics capable of distinguishing precise drumming from noisy motion. In this paper, we introduce a generative diffusion framework featuring a dual-objective loss function that decouples skeletal integrity from drumstick precision, thus enabling centimeter-level stick precision without sacrificing natural body dynamics. Additionally, leveraging our own dataset and data augmentation strategy, the model generalizes to non-curated, in-the-wild audio. To rigorously evaluate performance, we propose two novel metrics: an impact-to-target distance to quantify spatial precision and an audio-motion correlation score to assess temporal alignment. Our quantitative analysis and user studies demonstrate that our system generates high-quality motion that is often indistinguishable from ground-truth performances.

cs.CV↗

arXiv2026

ImageRAG: Dynamic Image Retrieval for Reference-Guided Image Generation

Diffusion models enable high-quality and diverse visual content synthesis. However, they struggle to generate rare or unseen concepts. To address this challenge, we explore the usage of Retrieval-Augmented Generation (RAG) with image generation models. We propose ImageRAG, a method that dynamically retrieves relevant images based on a given text prompt, and uses them as context to guide the generation process. Prior approaches that used retrieved images to improve generation, trained models specifically for retrieval-based generation. In contrast, ImageRAG leverages the capabilities of existing image conditioning models, and does not require RAG-specific training. Our approach is highly adaptable and can be applied across different model types, showing significant improvement in generating rare and fine-grained concepts using different base models. Our project page is available at: https://rotem-shalev.github.io/ImageRAG

cs.CV↗

arXiv2025

CARLoS: Retrieval via Concise Assessment Representation of LoRAs at Scale

The rapid proliferation of generative components, such as LoRAs, has created a vast but unstructured ecosystem. Existing discovery methods depend on unreliable user descriptions or biased popularity metrics, hindering usability. We present CARLoS, a large-scale framework for characterizing LoRAs without requiring additional metadata. Analyzing over 650 LoRAs, we employ them in image generation over a variety of prompts and seeds, as a credible way to assess their behavior. Using CLIP embeddings and their difference to a base-model generation, we concisely define a three-part representation: Directions, defining semantic shift; Strength, quantifying the significance of the effect; and Consistency, quantifying how stable the effect is. Using these representations, we develop an efficient retrieval framework that semantically matches textual queries to relevant LoRAs while filtering overly strong or unstable ones, outperforming textual baselines in automated and human evaluations. While retrieval is our primary focus, the same representation also supports analyses linking Strength and Consistency to legal notions of substantiality and volition, key considerations in copyright, positioning CARLoS as a practical system with broader relevance for LoRA analysis.

cs.AI↗

arXiv2025

Count The Notes: Histogram-Based Supervision for Automatic Music Transcription

Automatic Music Transcription (AMT) converts audio recordings into symbolic musical representations. Training deep neural networks (DNNs) for AMT typically requires strongly aligned training pairs with precise frame-level annotations. Since creating such datasets is costly and impractical for many musical contexts, weakly aligned approaches using segment-level annotations have gained traction. However, existing methods often rely on Dynamic Time Warping (DTW) or soft alignment loss functions, both of which still require local semantic correspondences, making them error-prone and computationally expensive. In this article, we introduce CountEM, a novel AMT framework that eliminates the need for explicit local alignment by leveraging note event histograms as supervision, enabling lighter computations and greater flexibility. Using an Expectation-Maximization (EM) approach, CountEM iteratively refines predictions based solely on note occurrence counts, significantly reducing annotation efforts while maintaining high transcription accuracy. Experiments on piano, guitar, and multi-instrument datasets demonstrate that CountEM matches or surpasses existing weakly supervised methods, improving AMT's robustness, scalability, and efficiency. Our project page is available at https://yoni-yaffe.github.io/count-the-notes.

cs.SD↗

arXiv2025

HOIDiNi: Human-Object Interaction through Diffusion Noise Optimization

We present HOIDiNi, a text-driven diffusion framework for synthesizing realistic and plausible human-object interaction (HOI). HOI generation is extremely challenging since it induces strict contact accuracies alongside a diverse motion manifold. While current literature trades off between realism and physical correctness, HOIDiNi optimizes directly in the noise space of a pretrained diffusion model using Diffusion Noise Optimization (DNO), achieving both. This is made feasible thanks to our observation that the problem can be separated into two phases: an object-centric phase, primarily making discrete choices of hand-object contact locations, and a human-centric phase that refines the full-body motion to realize this blueprint. This structured approach allows for precise hand-object contact without compromising motion naturalness. Quantitative, qualitative, and subjective evaluations on the GRAB dataset alone clearly indicate HOIDiNi outperforms prior works and baselines in contact accuracy, physical validity, and overall quality. Our results demonstrate the ability to generate complex, controllable interactions, including grasping, placing, and full-body coordination, driven solely by textual prompts. https://hoidini.github.io.

cs.CV↗

arXiv2025

Attention (as Discrete-Time Markov) Chains

We introduce a new interpretation of the attention matrix as a discrete-time Markov chain. Our interpretation sheds light on common operations involving attention scores such as selection, summation, and averaging in a unified framework. It further extends them by considering indirect attention, propagated through the Markov chain, as opposed to previous studies that only model immediate effects. Our key observation is that tokens linked to semantically similar regions form metastable states, i.e., regions where attention tends to concentrate, while noisy attention scores dissipate. Metastable states and their prevalence can be easily computed through simple matrix multiplication and eigenanalysis, respectively. Using these lightweight tools, we demonstrate state-of-the-art zero-shot segmentation. Lastly, we define TokenRank -- the steady state vector of the Markov chain, which measures global token importance. We show that TokenRank enhances unconditional image generation, improving both quality (IS) and diversity (FID), and can also be incorporated into existing segmentation techniques to improve their performance over existing benchmarks. We believe our framework offers a fresh view of how tokens are being attended in modern visual transformers.

cs.CV↗

arXiv2025

PractiLight: Practical Light Control Using Foundational Diffusion Models

Light control in generated images is a difficult task, posing specific challenges, spanning over the entire image and frequency spectrum. Most approaches tackle this problem by training on extensive yet domain-specific datasets, limiting the inherent generalization and applicability of the foundational backbones used. Instead, PractiLight is a practical approach, effectively leveraging foundational understanding of recent generative models for the task. Our key insight is that lighting relationships in an image are similar in nature to token interaction in self-attention layers, and hence are best represented there. Based on this and other analyses regarding the importance of early diffusion iterations, PractiLight trains a lightweight LoRA regressor to produce the direct irradiance map for a given image, using a small set of training images. We then employ this regressor to incorporate the desired lighting into the generation process of another image using Classifier Guidance. This careful design generalizes well to diverse conditions and image domains. We demonstrate state-of-the-art performance in terms of quality and control with proven parameter and data efficiency compared to leading works over a wide variety of scenes types. We hope this work affirms that image lighting can feasibly be controlled by tapping into foundational knowledge, enabling practical and general relighting.

cs.CV↗

arXiv2025

Dance Like a Chicken: Low-Rank Stylization for Human Motion Diffusion

Text-to-motion generative models span a wide range of 3D human actions but struggle with nuanced stylistic attributes such as a "Chicken" style. Due to the scarcity of style-specific data, existing approaches pull the generative prior towards a reference style, which often results in out-of-distribution low quality generations. In this work, we introduce LoRA-MDM, a lightweight framework for motion stylization that generalizes to complex actions while maintaining editability. Our key insight is that adapting the generative prior to include the style, while preserving its overall distribution, is more effective than modifying each individual motion during generation. Building on this idea, LoRA-MDM learns to adapt the prior to include the reference style using only a few samples. The style can then be used in the context of different textual prompts for generation. The low-rank adaptation shifts the motion manifold in a semantically meaningful way, enabling realistic style infusion even for actions not present in the reference samples. Moreover, preserving the distribution structure enables advanced operations such as style blending and motion editing. We compare LoRA-MDM to state-of-the-art stylized motion generation methods and demonstrate a favorable balance between text fidelity and style consistency.

cs.CV↗

arXiv2025

AnyTop: Character Animation Diffusion with Any Topology

Generating motion for arbitrary skeletons is a longstanding challenge in computer graphics, remaining largely unexplored due to the scarcity of diverse datasets and the irregular nature of the data. In this work, we introduce AnyTop, a diffusion model that generates motions for diverse characters with distinct motion dynamics, using only their skeletal structure as input. Our work features a transformer-based denoising network, tailored for arbitrary skeleton learning, integrating topology information into the traditional attention mechanism. Additionally, by incorporating textual joint descriptions into the latent feature representation, AnyTop learns semantic correspondences between joints across diverse skeletons. Our evaluation demonstrates that AnyTop generalizes well, even with as few as three training examples per topology, and can produce motions for unseen skeletons as well. Furthermore, our model's latent space is highly informative, enabling downstream tasks such as joint correspondence, temporal segmentation and motion editing. Our webpage, https://anytop2025.github.io/Anytop-page, includes links to videos and code.

cs.GR↗

arXiv2025

V-LASIK: Consistent Glasses-Removal from Videos Using Synthetic Data

Diffusion-based generative models have recently shown remarkable image and video editing capabilities. However, local video editing, particularly removal of small attributes like glasses, remains a challenge. Existing methods either alter the videos excessively, generate unrealistic artifacts, or fail to perform the requested edit consistently throughout the video. In this work, we focus on consistent and identity-preserving removal of glasses in videos, using it as a case study for consistent local attribute removal in videos. Due to the lack of paired data, we adopt a weakly supervised approach and generate synthetic imperfect data, using an adjusted pretrained diffusion model. We show that despite data imperfection, by learning from our generated data and leveraging the prior of pretrained diffusion models, our model is able to perform the desired edit consistently while preserving the original video content. Furthermore, we exemplify the generalization ability of our method to other local video editing tasks by applying it successfully to facial sticker-removal. Our approach demonstrates significant improvement over existing methods, showcasing the potential of leveraging synthetic data and strong video priors for local video editing tasks.

cs.CV↗

arXiv2024

Instant3dit: Multiview Inpainting for Fast Editing of 3D Objects

We propose a generative technique to edit 3D shapes, represented as meshes, NeRFs, or Gaussian Splats, in approximately 3 seconds, without the need for running an SDS type of optimization. Our key insight is to cast 3D editing as a multiview image inpainting problem, as this representation is generic and can be mapped back to any 3D representation using the bank of available Large Reconstruction Models. We explore different fine-tuning strategies to obtain both multiview generation and inpainting capabilities within the same diffusion model. In particular, the design of the inpainting mask is an important factor of training an inpainting model, and we propose several masking strategies to mimic the types of edits a user would perform on a 3D shape. Our approach takes 3D generative editing from hours to seconds and produces higher-quality results compared to previous works.

cs.CV↗

arXiv2024

MagicClay: Sculpting Meshes With Generative Neural Fields

The recent developments in neural fields have brought phenomenal capabilities to the field of shape generation, but they lack crucial properties, such as incremental control - a fundamental requirement for artistic work. Triangular meshes, on the other hand, are the representation of choice for most geometry related tasks, offering efficiency and intuitive control, but do not lend themselves to neural optimization. To support downstream tasks, previous art typically proposes a two-step approach, where first a shape is generated using neural fields, and then a mesh is extracted for further processing. Instead, in this paper we introduce a hybrid approach that maintains both a mesh and a Signed Distance Field (SDF) representations consistently. Using this representation, we introduce MagicClay - an artist friendly tool for sculpting regions of a mesh according to textual prompts while keeping other regions untouched. Our framework carefully and efficiently balances consistency between the representations and regularizations in every step of the shape optimization; Relying on the mesh representation, we show how to render the SDF at higher resolutions and faster. In addition, we employ recent work in differentiable mesh reconstruction to adaptively allocate triangles in the mesh where required, as indicated by the SDF. Using an implemented prototype, we demonstrate superior generated geometry compared to the state-of-the-art, and novel consistent control, allowing sequential prompt-based edits to the same mesh for the first time.

cs.GR↗

arXiv2024

CLoSD: Closing the Loop between Simulation and Diffusion for multi-task character control

Motion diffusion models and Reinforcement Learning (RL) based control for physics-based simulations have complementary strengths for human motion generation. The former is capable of generating a wide variety of motions, adhering to intuitive control such as text, while the latter offers physically plausible motion and direct interaction with the environment. In this work, we present a method that combines their respective strengths. CLoSD is a text-driven RL physics-based controller, guided by diffusion generation for various tasks. Our key insight is that motion diffusion can serve as an on-the-fly universal planner for a robust RL controller. To this end, CLoSD maintains a closed-loop interaction between two modules -- a Diffusion Planner (DiP), and a tracking controller. DiP is a fast-responding autoregressive diffusion model, controlled by textual prompts and target locations, and the controller is a simple and robust motion imitator that continuously receives motion plans from DiP and provides feedback from the environment. CLoSD is capable of seamlessly performing a sequence of different tasks, including navigation to a goal location, striking an object with a hand or foot as specified in a text prompt, sitting down, and getting up. https://guytevet.github.io/CLoSD-page/

cs.CV↗

arXiv2024

ComfyGen: Prompt-Adaptive Workflows for Text-to-Image Generation

The practical use of text-to-image generation has evolved from simple, monolithic models to complex workflows that combine multiple specialized components. While workflow-based approaches can lead to improved image quality, crafting effective workflows requires significant expertise, owing to the large number of available components, their complex inter-dependence, and their dependence on the generation prompt. Here, we introduce the novel task of prompt-adaptive workflow generation, where the goal is to automatically tailor a workflow to each user prompt. We propose two LLM-based approaches to tackle this task: a tuning-based method that learns from user-preference data, and a training-free method that uses the LLM to select existing flows. Both approaches lead to improved image quality when compared to monolithic models or generic, prompt-independent workflows. Our work shows that prompt-dependent flow prediction offers a new pathway to improving text-to-image generation quality, complementing existing research directions in the field.

cs.CV↗

arXiv2024

Casper DPM: Cascaded Perceptual Dynamic Projection Mapping onto Hands

We present a technique for dynamically projecting 3D content onto human hands with short perceived motion-to-photon latency. Computing the pose and shape of human hands accurately and quickly is a challenging task due to their articulated and deformable nature. We combine a slower 3D coarse estimation of the hand pose with high speed 2D correction steps which improve the alignment of the projection to the hands, increase the projected surface area, and reduce perceived latency. Since our approach leverages a full 3D reconstruction of the hands, any arbitrary texture or reasonably performant effect can be applied, which was not possible before. We conducted two user studies to assess the benefits of using our method. The results show subjects are less sensitive to latency artifacts and perform faster and with more ease a given associated task over the naive approach of directly projecting rendered frames from the 3D pose estimation. We demonstrate several novel use cases and applications.

cs.GR↗

arXiv2024

Masked Extended Attention for Zero-Shot Virtual Try-On In The Wild

Virtual Try-On (VTON) is a highly active line of research, with increasing demand. It aims to replace a piece of garment in an image with one from another, while preserving person and garment characteristics as well as image fidelity. Current literature takes a supervised approach for the task, impairing generalization and imposing heavy computation. In this paper, we present a novel zero-shot training-free method for inpainting a clothing garment by reference. Our approach employs the prior of a diffusion model with no additional training, fully leveraging its native generalization capabilities. The method employs extended attention to transfer image information from reference to target images, overcoming two significant challenges. We first initially warp the reference garment over the target human using deep features, alleviating "texture sticking". We then leverage the extended attention mechanism with careful masking, eliminating leakage of reference background and unwanted influence. Through a user study, qualitative, and quantitative comparison to state-of-the-art approaches, we demonstrate superior image quality and garment preservation compared unseen clothing pieces or human figures.

cs.CV↗

arXiv2024

Monkey See, Monkey Do: Harnessing Self-attention in Motion Diffusion for Zero-shot Motion Transfer

Given the remarkable results of motion synthesis with diffusion models, a natural question arises: how can we effectively leverage these models for motion editing? Existing diffusion-based motion editing methods overlook the profound potential of the prior embedded within the weights of pre-trained models, which enables manipulating the latent feature space; hence, they primarily center on handling the motion space. In this work, we explore the attention mechanism of pre-trained motion diffusion models. We uncover the roles and interactions of attention elements in capturing and representing intricate human motion patterns, and carefully integrate these elements to transfer a leader motion to a follower one while maintaining the nuanced characteristics of the follower, resulting in zero-shot motion transfer. Editing features associated with selected motions allows us to confront a challenge observed in prior motion diffusion approaches, which use general directives (e.g., text, music) for editing, ultimately failing to convey subtle nuances effectively. Our work is inspired by how a monkey closely imitates what it sees while maintaining its unique motion patterns; hence we call it Monkey See, Monkey Do, and dub it MoMo. Employing our technique enables accomplishing tasks such as synthesizing out-of-distribution motions, style transfer, and spatial editing. Furthermore, diffusion inversion is seldom employed for motions; as a result, editing efforts focus on generated motions, limiting the editability of real ones. MoMo harnesses motion inversion, extending its application to both real and generated motions. Experimental results show the advantage of our approach over the current art. In particular, unlike methods tailored for specific applications through training, our approach is applied at inference time, requiring no training. Our webpage is at https://monkeyseedocg.github.io.

cs.CV↗

arXiv2024

LCM-Lookahead for Encoder-based Text-to-Image Personalization

Recent advancements in diffusion models have introduced fast sampling methods that can effectively produce high-quality images in just one or a few denoising steps. Interestingly, when these are distilled from existing diffusion models, they often maintain alignment with the original model, retaining similar outputs for similar prompts and seeds. These properties present opportunities to leverage fast sampling methods as a shortcut-mechanism, using them to create a preview of denoised outputs through which we can backpropagate image-space losses. In this work, we explore the potential of using such shortcut-mechanisms to guide the personalization of text-to-image models to specific facial identities. We focus on encoder-based personalization approaches, and demonstrate that by tuning them with a lookahead identity loss, we can achieve higher identity fidelity, without sacrificing layout diversity or prompt alignment. We further explore the use of attention sharing mechanisms and consistent data generation for the task of personalization, and find that encoder training can benefit from both.

cs.CV↗