arXiv subjects

Yandong Li

Publications and source records attributed to Yandong Li.

At least 19 recordsLinked to original sources

arXiv2026

What should a linear optical frontend compute? Assessing the role of meta-optics, nonlocality, and coherence in hybrid inference systems

Hybrid inference systems that pair an optical frontend with a digital backend offer a route to offload computation to the physical layer. Yet what the optics should compute, and when this is beneficial, has remained unclear. Here, we show that, for classification tasks, a well-designed optical frontend reshapes the statistics of the sensor intensity readout to improve class separability, quantified by the Bhattacharyya distance. This training-free metric predicts downstream accuracy and reveals that much of the discriminative information resides in inter-pixel correlations. We then identify the roles of coherence and different forms of nonlocality. Because the sensor measures intensity, a linear frontend produces features that are quadratic in the input field; however, only nonlocal, coherent optical systems can exploit the associated information. Such systems can yield significant performance gains, surpassing the best trained linear preprocessor. These results provide physical insights and new design principles for optimal optical--electronic inference systems.

physics.optics↗

arXiv2026

Image Generators are Generalist Vision Learners

Recent works show that image and video generators exhibit zero-shot visual understanding behaviors, in a way reminiscent of how LLMs develop emergent capabilities of language understanding and reasoning from generative pretraining. While it has long been conjectured that the ability to create visual content implies an ability to understand it, there has been limited evidence that generative vision models have developed strong understanding capabilities. In this work, we demonstrate that image generation training serves a role similar to LLM pretraining, and lets models learn powerful and general visual representations that enable SOTA performance on various vision tasks. We introduce Vision Banana, a generalist model built by instruction-tuning Nano Banana Pro (NBP) on a mixture of its original training data alongside a small amount of vision task data. By parameterizing the output space of vision tasks as RGB images, we seamlessly reframe perception as image generation. Our generalist model, Vision Banana, achieves SOTA results on a variety of vision tasks involving both 2D and 3D understanding, beating or rivaling zero-shot domain-specialists, including Segment Anything Model 3 on segmentation tasks, and the Depth Anything series on metric depth estimation. We show that these results can be achieved with lightweight instruction-tuning without sacrificing the base model's image generation capabilities. The superior results suggest that image generation pretraining is a generalist vision learner. It also shows that image generation serves as a unified and universal interface for vision tasks, similar to text generation's role in language understanding and reasoning. We could be witnessing a major paradigm shift for computer vision, where generative vision pretraining takes a central role in building Foundational Vision Models for both generation and understanding.

cs.CV↗

arXiv2025

Epsilon-VAE: Denoising as Visual Decoding

In generative modeling, tokenization simplifies complex data into compact, structured representations, creating a more efficient, learnable space. For high-dimensional visual data, it reduces redundancy and emphasizes key features for high-quality generation. Current visual tokenization methods rely on a traditional autoencoder framework, where the encoder compresses data into latent representations, and the decoder reconstructs the original input. In this work, we offer a new perspective by proposing denoising as decoding, shifting from single-step reconstruction to iterative refinement. Specifically, we replace the decoder with a diffusion process that iteratively refines noise to recover the original image, guided by the latents provided by the encoder. We evaluate our approach by assessing both reconstruction (rFID) and generation quality (FID), comparing it to state-of-the-art autoencoding approaches. By adopting iterative reconstruction through diffusion, our autoencoder, namely Epsilon-VAE, achieves high reconstruction quality, which in turn enhances downstream generation quality by 22% at the same compression rates or provides 2.3x inference speedup through increasing compression rates. We hope this work offers new insights into integrating iterative generation and autoencoding for improved compression and generation.

cs.CV↗

arXiv2025

The Spatial Complexity of Optical Computing and How to Reduce It

Similar to algorithms, which consume time and memory to run, hardware requires resources to function. For devices processing physical waves, implementing operations needs sufficient "space," as dictated by wave physics. How much space is needed to perform a certain function is a fundamental question in optics, with recent research addressing it for given mathematical operations, but not for more general computing tasks, e.g., classification. Inspired by computational complexity theory, we study the "spatial complexity" of optical computing systems in terms of scaling laws - specifically, how their physical dimensions must scale as the dimension of the mathematical operation increases - and propose a new paradigm for designing optical computing systems: space-efficient neuromorphic optics, based on structural sparsity constraints and neural pruning methods motivated by wave physics (notably, the concept of "overlapping nonlocality"). On two mainstream platforms, free-space optics and on-chip integrated photonics, our methods demonstrate substantial size reductions (to 1%-10% the size of conventional designs) with minimal compromise on performance. Our theoretical and computational results reveal a trend of diminishing returns on accuracy as structure dimensions increase, providing a new perspective for interpreting and approaching the ultimate limits of optical computing - a balanced trade-off between device size and accuracy.

physics.optics↗

arXiv2025

Inference-Time Scaling for Diffusion Models beyond Scaling Denoising Steps

Generative models have made significant impacts across various domains, largely due to their ability to scale during training by increasing data, computational resources, and model size, a phenomenon characterized by the scaling laws. Recent research has begun to explore inference-time scaling behavior in Large Language Models (LLMs), revealing how performance can further improve with additional computation during inference. Unlike LLMs, diffusion models inherently possess the flexibility to adjust inference-time computation via the number of denoising steps, although the performance gains typically flatten after a few dozen. In this work, we explore the inference-time scaling behavior of diffusion models beyond increasing denoising steps and investigate how the generation performance can further improve with increased computation. Specifically, we consider a search problem aimed at identifying better noises for the diffusion sampling process. We structure the design space along two axes: the verifiers used to provide feedback, and the algorithms used to find better noise candidates. Through extensive experiments on class-conditioned and text-conditioned image generation benchmarks, our findings reveal that increasing inference-time compute leads to substantial improvements in the quality of samples generated by diffusion models, and with the complicated nature of images, combinations of the components in the framework can be specifically chosen to conform with different application scenario.

cs.CV↗

arXiv2025

Broadband localization of light at the termination of a topological photonic waveguide

Localized optical field enhancement enables strong light-matter interactions necessary for efficient manipulation and sensing of light. Specifically, tunable broadband energy localization in nanoscale hotspots offers a wide range of applications in nanophotonics and quantum optics. We experimentally demonstrate a novel principle for the local enhancement of electromagnetic fields, based on strong suppression of backscattering. This is achieved at a designed termination of a topologically non-trivial waveguide that nearly preserves the valley degree of freedom. The symmetry origin of the valley degree of freedom prevents edge states to undergo intervalley scattering at waveguide discontinuities that obey the symmetry of the crystal. Using near-field microscopy, we reveal that this can lead to strong confinement of light at the termination of a topological photonic waveguide, even without breaking time-reversal symmetry. We emphasize the importance of symmetry conservation by comparing different waveguide termination geometries, confirming that the origin of suppressed backscattering lies with the near-conservation of the valley degree of freedom, and show the broad bandwidth of the effect.

physics.optics↗

arXiv2024

Video Creation by Demonstration

We explore a novel video creation experience, namely Video Creation by Demonstration. Given a demonstration video and a context image from a different scene, we generate a physically plausible video that continues naturally from the context image and carries out the action concepts from the demonstration. To enable this capability, we present $δ$-Diffusion, a self-supervised training approach that learns from unlabeled videos by conditional future frame prediction. Unlike most existing video generation controls that are based on explicit signals, we adopts the form of implicit latent control for maximal flexibility and expressiveness required by general videos. By leveraging a video foundation model with an appearance bottleneck design on top, we extract action latents from demonstration videos for conditioning the generation process with minimal appearance leakage. Empirically, $δ$-Diffusion outperforms related baselines in terms of both human preference and large-scale machine evaluations, and demonstrates potentials towards interactive world simulation. Sampled video generation results are available at https://delta-diffusion.github.io/.

cs.CV↗

arXiv2024

OmnixR: Evaluating Omni-modality Language Models on Reasoning across Modalities

We introduce OmnixR, an evaluation suite designed to benchmark SoTA Omni-modality Language Models, such as GPT-4o and Gemini. Evaluating OLMs, which integrate multiple modalities such as text, vision, and audio, presents unique challenges. Particularly, the user message might often consist of multiple modalities, such that OLMs have to establish holistic understanding and reasoning across modalities to accomplish the task. Existing benchmarks are limited to single modality or dual-modality tasks, overlooking comprehensive multi-modal assessments of model reasoning. To address this, OmnixR offers two evaluation variants: (1)synthetic subset: a synthetic dataset generated automatically by translating text into multiple modalities--audio, images, video, and hybrids (Omnify). (2)realistic subset: a real-world dataset, manually curated and annotated by experts, for evaluating cross-modal reasoning in natural settings. OmnixR presents a unique evaluation towards assessing OLMs over a diverse mix of modalities, such as a question that involves video, audio, and text, providing a rigorous cross-modal reasoning testbed unlike any existing benchmarks. Our experiments find that all state-of-the-art OLMs struggle with OmnixR questions that require integrating information from multiple modalities to answer. Further analysis highlights differences in reasoning behavior, underscoring the challenges of omni-modal AI alignment.

cs.AI↗

arXiv2024

What's in a Name? Beyond Class Indices for Image Recognition

Existing machine learning models demonstrate excellent performance in image object recognition after training on a large-scale dataset under full supervision. However, these models only learn to map an image to a predefined class index, without revealing the actual semantic meaning of the object in the image. In contrast, vision-language models like CLIP are able to assign semantic class names to unseen objects in a 'zero-shot' manner, though they are once again provided a pre-defined set of candidate names at test-time. In this paper, we reconsider the recognition problem and task a vision-language model with assigning class names to images given only a large (essentially unconstrained) vocabulary of categories as prior information. We leverage non-parametric methods to establish meaningful relationships between images, allowing the model to automatically narrow down the pool of candidate names. Our proposed approach entails iteratively clustering the data and employing a voting mechanism to determine the most suitable class names. Additionally, we investigate the potential of incorporating additional textual features to enhance clustering performance. To achieve this, we employ the CLIP vision and text encoders to retrieve relevant texts from an external database, which can provide supplementary semantic information to inform the clustering process. Furthermore, we tackle this problem both in unsupervised and partially supervised settings, as well as with a coarse-grained and fine-grained search space as the unconstrained dictionary. Remarkably, our method leads to a roughly 50% improvement over the baseline on ImageNet in the unsupervised setting.

cs.CV↗

arXiv2024

Perturbing Attention Gives You More Bang for the Buck: Subtle Imaging Perturbations That Efficiently Fool Customized Diffusion Models

Diffusion models (DMs) embark a new era of generative modeling and offer more opportunities for efficient generating high-quality and realistic data samples. However, their widespread use has also brought forth new challenges in model security, which motivates the creation of more effective adversarial attackers on DMs to understand its vulnerability. We propose CAAT, a simple but generic and efficient approach that does not require costly training to effectively fool latent diffusion models (LDMs). The approach is based on the observation that cross-attention layers exhibits higher sensitivity to gradient change, allowing for leveraging subtle perturbations on published images to significantly corrupt the generated images. We show that a subtle perturbation on an image can significantly impact the cross-attention layers, thus changing the mapping between text and image during the fine-tuning of customized diffusion models. Extensive experiments demonstrate that CAAT is compatible with diverse diffusion models and outperforms baseline attack methods in a more effective (more noise) and efficient (twice as fast as Anti-DreamBooth and Mist) manner.

cs.CV↗

arXiv2024

Topological Directional Coupler

Interferometers and beam splitters are fundamental building blocks for photonic neuromorphic and quantum computing machinery. In waveguide-based photonic integrated circuits, beam-splitting is achieved with directional couplers that rely on transition regions where the waveguides are adiabatically bent to suppress back-reflection. We present a novel, compact approach to introducing guided mode coupling. By leveraging multimodal domain walls between microwave topological photonic crystals, we use the photonic-spin-conservation to suppress back-reflection while relaxing the topological protection of the valley degree of freedom to implement tunable beam splitting. Rapid advancements in chip-scale topological photonics suggest that the proposed simultaneous utilization of multiple topological degrees of freedom could benefit the development of novel photonic computing platforms.

cond-mat.mes-hall↗

arXiv2024

RoboDreamer: Learning Compositional World Models for Robot Imagination

Text-to-video models have demonstrated substantial potential in robotic decision-making, enabling the imagination of realistic plans of future actions as well as accurate environment simulation. However, one major issue in such models is generalization -- models are limited to synthesizing videos subject to language instructions similar to those seen at training time. This is heavily limiting in decision-making, where we seek a powerful world model to synthesize plans of unseen combinations of objects and actions in order to solve previously unseen tasks in new environments. To resolve this issue, we introduce RoboDreamer, an innovative approach for learning a compositional world model by factorizing the video generation. We leverage the natural compositionality of language to parse instructions into a set of lower-level primitives, which we condition a set of models on to generate videos. We illustrate how this factorization naturally enables compositional generalization, by allowing us to formulate a new natural language instruction as a combination of previously seen components. We further show how such a factorization enables us to add additional multimodal goals, allowing us to specify a video we wish to generate given both natural language instructions and a goal image. Our approach can successfully synthesize video plans on unseen goals in the RT-X, enables successful robot execution in simulation, and substantially outperforms monolithic baseline approaches to video generation.

cs.RO↗

arXiv2024

Instruct-Imagen: Image Generation with Multi-modal Instruction

This paper presents instruct-imagen, a model that tackles heterogeneous image generation tasks and generalizes across unseen tasks. We introduce *multi-modal instruction* for image generation, a task representation articulating a range of generation intents with precision. It uses natural language to amalgamate disparate modalities (e.g., text, edge, style, subject, etc.), such that abundant generation intents can be standardized in a uniform format. We then build instruct-imagen by fine-tuning a pre-trained text-to-image diffusion model with a two-stage framework. First, we adapt the model using the retrieval-augmented training, to enhance model's capabilities to ground its generation on external multimodal context. Subsequently, we fine-tune the adapted model on diverse image generation tasks that requires vision-language understanding (e.g., subject-driven generation, etc.), each paired with a multi-modal instruction encapsulating the task's essence. Human evaluation on various image generation datasets reveals that instruct-imagen matches or surpasses prior task-specific models in-domain and demonstrates promising generalization to unseen and more complex tasks.

cs.CV↗

arXiv2023

DreamInpainter: Text-Guided Subject-Driven Image Inpainting with Diffusion Models

This study introduces Text-Guided Subject-Driven Image Inpainting, a novel task that combines text and exemplar images for image inpainting. While both text and exemplar images have been used independently in previous efforts, their combined utilization remains unexplored. Simultaneously accommodating both conditions poses a significant challenge due to the inherent balance required between editability and subject fidelity. To tackle this challenge, we propose a two-step approach DreamInpainter. First, we compute dense subject features to ensure accurate subject replication. Then, we employ a discriminative token selection module to eliminate redundant subject details, preserving the subject's identity while allowing changes according to other conditions such as mask shape and text prompts. Additionally, we introduce a decoupling regularization technique to enhance text control in the presence of exemplar images. Our extensive experiments demonstrate the superior performance of our method in terms of visual quality, identity preservation, and text control, showcasing its effectiveness in the context of text-guided subject-driven image inpainting.

cs.CV↗

arXiv2023

Deep Subwavelength Topological Edge State in a Hyperbolic Medium

Topological nanophotonics presents the potential for cutting-edge photonic systems, with a core aim revolving around the emergence of topological edge states. These states are primed to propagate robustly while embracing deep subwavelength confinement that defies diffraction limits. Such attributes make them particularly appealing for nanoscale applications, where achieving these elusive states has remained challenging. We unveil the first experimental proof of deep subwavelength topological edge states by implementing periodic modulation of hyperbolic phonon polaritons within a Van der Waals heterostructure. This finding represents a significant milestone in the field of nanophotonics, and it can be directly extended to and hybridized with other Van der Waals materials in various applications. The extensive scope for material substitution facilitates broadened operational frequency ranges, streamlined integration of diverse polaritonic materials, and compatibility with electronic and excitonic systems.

physics.optics↗

arXiv2023

Subject-driven Text-to-Image Generation via Apprenticeship Learning

Recent text-to-image generation models like DreamBooth have made remarkable progress in generating highly customized images of a target subject, by fine-tuning an ``expert model'' for a given subject from a few examples. However, this process is expensive, since a new expert model must be learned for each subject. In this paper, we present SuTI, a Subject-driven Text-to-Image generator that replaces subject-specific fine tuning with in-context learning. Given a few demonstrations of a new subject, SuTI can instantly generate novel renditions of the subject in different scenes, without any subject-specific optimization. SuTI is powered by apprenticeship learning, where a single apprentice model is learned from data generated by a massive number of subject-specific expert models. Specifically, we mine millions of image clusters from the Internet, each centered around a specific visual subject. We adopt these clusters to train a massive number of expert models, each specializing in a different subject. The apprentice model SuTI then learns to imitate the behavior of these fine-tuned experts. SuTI can generate high-quality and customized subject-specific images 20x faster than optimization-based SoTA methods. On the challenging DreamBench and DreamBench-v2, our human evaluation shows that SuTI significantly outperforms existing models like InstructPix2Pix, Textual Inversion, Imagic, Prompt2Prompt, Re-Imagen and DreamBooth, especially on the subject and text alignment aspects.

cs.CV↗

arXiv2023

MUG: Multi-human Graph Network for 3D Mesh Reconstruction from 2D Pose

Reconstructing multi-human body mesh from a single monocular image is an important but challenging computer vision problem. In addition to the individual body mesh models, we need to estimate relative 3D positions among subjects to generate a coherent representation. In this work, through a single graph neural network, named MUG (Multi-hUman Graph network), we construct coherent multi-human meshes using only multi-human 2D pose as input. Compared with existing methods, which adopt a detection-style pipeline (i.e., extracting image features and then locating human instances and recovering body meshes from that) and suffer from the significant domain gap between lab-collected training datasets and in-the-wild testing datasets, our method benefits from the 2D pose which has a relatively consistent geometric property across datasets. Our method works like the following: First, to model the multi-human environment, it processes multi-human 2D poses and builds a novel heterogeneous graph, where nodes from different people and within one person are connected to capture inter-human interactions and draw the body geometry (i.e., skeleton and mesh structure). Second, it employs a dual-branch graph neural network structure -- one for predicting inter-human depth relation and the other one for predicting root-joint-relative mesh coordinates. Finally, the entire multi-human 3D meshes are constructed by combining the output from both branches. Extensive experiments demonstrate that MUG outperforms previous multi-human mesh estimation methods on standard 3D human benchmarks -- Panoptic, MuPoTS-3D and 3DPW.

cs.CV↗

arXiv2023

Identity Encoder for Personalized Diffusion

Many applications can benefit from personalized image generation models, including image enhancement, video conferences, just to name a few. Existing works achieved personalization by fine-tuning one model for each person. While being successful, this approach incurs additional computation and storage overhead for each new identity. Furthermore, it usually expects tens or hundreds of examples per identity to achieve the best performance. To overcome these challenges, we propose an encoder-based approach for personalization. We learn an identity encoder which can extract an identity representation from a set of reference images of a subject, together with a diffusion generator that can generate new images of the subject conditioned on the identity representation. Once being trained, the model can be used to generate images of arbitrary identities given a few examples even if the model hasn't been trained on the identity. Our approach greatly reduces the overhead for personalized image generation and is more applicable in many potential applications. Empirical results show that our approach consistently outperforms existing fine-tuning based approach in both image generation and reconstruction, and the outputs is preferred by users more than 95% of the time compared with the best performing baseline.

cs.CV↗