Search

Text-guided diffusion models have significantly advanced image editing, enabling high-quality and diverse modifications driven by text prompts. However, effective editing requires inverting the source image into a latent space, a process often hindered by prediction errors inherent in DDIM inversion. These errors accumulate during the diffusion process, resulting in inferior content preservation and edit fidelity, especially with conditional inputs. We address these challenges by investigating the primary contributors to error accumulation in DDIM inversion and identify the singularity problem in traditional noise schedules as a key issue. To resolve this, we introduce the Logistic Schedule, a novel noise schedule designed to eliminate singularities, improve inversion stability, and provide a better noise space for image editing. This schedule reduces noise prediction errors, enabling more faithful editing that preserves the original content of the source image. Our approach requires no additional retraining and is compatible with various existing editing methods. Experiments across eight editing tasks demonstrate the Logistic Schedule's superior performance in content preservation and edit fidelity compared to traditional noise schedules, highlighting its adaptability and effectiveness., Comment: Accepted in NeurIPS 2024

Recent advancements have shown promise in applying traditional Semi-Supervised Learning strategies to the task of Generalized Category Discovery (GCD). Typically, this involves a teacher-student framework in which the teacher imparts knowledge to the student to classify categories, even in the absence of explicit labels. Nevertheless, GCD presents unique challenges, particularly the absence of priors for new classes, which can lead to the teacher's misguidance and unsynchronized learning with the student, culminating in suboptimal outcomes. In our work, we delve into why traditional teacher-student designs falter in open-world generalized category discovery as compared to their success in closed-world semi-supervised learning. We identify inconsistent pattern learning across attention layers as the crux of this issue and introduce FlipClass, a method that dynamically updates the teacher to align with the student's attention, instead of maintaining a static teacher reference. Our teacher-student attention alignment strategy refines the teacher's focus based on student feedback from an energy perspective, promoting consistent pattern recognition and synchronized learning across old and new classes. Extensive experiments on a spectrum of benchmarks affirm that FlipClass significantly surpasses contemporary GCD methods, establishing new standards for the field., Comment: Accepted by NeurIPS 2024 (Oral)

Hyperspectral imaging has been widely used for spectral and spatial identification of target molecules, yet often contaminated by sophisticated noise. Current denoising methods generally rely on independent and identically distributed noise statistics, showing corrupted performance for non-independent noise removal. Here, we demonstrate Self-supervised PErmutation Noise2noise Denoising (SPEND), a deep learning denoising architecture tailor-made for removing non-independent noise from a single hyperspectral image stack. We utilize hyperspectral stimulated Raman scattering and mid-infrared photothermal microscopy as the testbeds, where the noise is spatially correlated and spectrally varied. Based on single hyperspectral images, SPEND permutates odd and even spectral frames to generate two stacks with identical noise properties, and uses the pairs for efficient self-supervised noise-to-noise training. SPEND achieved an 8-fold signal-to-noise improvement without having access to the ground truth data. SPEND enabled accurate mapping of low concentration biomolecules in both fingerprint and silent regions, demonstrating its robustness in sophisticated cellular environments.

Text-to-image diffusion models significantly enhance the efficiency of artistic creation with high-fidelity image generation. However, in typical application scenarios like comic book production, they can neither place each subject into its expected spot nor maintain the consistent appearance of each subject across images. For these issues, we pioneer a novel task, Layout-to-Consistent-Image (L2CI) generation, which produces consistent and compositional images in accordance with the given layout conditions and text prompts. To accomplish this challenging task, we present a new formalization of dual energy guidance with optimization in a dual semantic-latent space and thus propose a training-free pipeline, SpotActor, which features a layout-conditioned backward update stage and a consistent forward sampling stage. In the backward stage, we innovate a nuanced layout energy function to mimic the attention activations with a sigmoid-like objective. While in the forward stage, we design Regional Interconnection Self-Attention (RISA) and Semantic Fusion Cross-Attention (SFCA) mechanisms that allow mutual interactions across images. To evaluate the performance, we present ActorBench, a specified benchmark with hundreds of reasonable prompt-box pairs stemming from object detection datasets. Comprehensive experiments are conducted to demonstrate the effectiveness of our method. The results prove that SpotActor fulfills the expectations of this task and showcases the potential for practical applications with superior layout alignment, subject consistency, prompt conformity and background diversity.

Knowledge-based Visual Question Answering (KVQA) requires both image and world knowledge to answer questions. Current methods first retrieve knowledge from the image and external knowledge base with the original complex question, then generate answers with Large Language Models (LLMs). However, since the original question contains complex elements that require knowledge from different sources, acquiring different kinds of knowledge in a coupled manner may confuse models and hinder them from retrieving precise knowledge. Furthermore, the ``forward-only'' answering process fails to explicitly capture the knowledge needs of LLMs, which can further hurt answering quality. To cope with the above limitations, we propose DKA: Disentangled Knowledge Acquisition from LLM feedback, a training-free framework that disentangles knowledge acquisition to avoid confusion and uses LLM's feedback to specify the required knowledge. Specifically, DKA requires LLMs to specify what knowledge they need to answer the question and decompose the original complex question into two simple sub-questions: Image-based sub-question and Knowledge-based sub-question. Then we use the two sub-questions to retrieve knowledge from the image and knowledge base, respectively. In this way, two knowledge acquisition models can focus on the content that corresponds to them and avoid disturbance of irrelevant elements in the original complex question, which can help to provide more precise knowledge and better align the knowledge needs of LLMs to yield correct answers. Experiments on benchmark datasets show that DKA significantly outperforms SOTA models. To facilitate future research, our data and code are available at \url{https://github.com/Lackel/DKA}., Comment: Pre-print

Diffusion models have marked a significant breakthrough in the synthesis of semantically coherent images. However, their extensive noise estimation networks and the iterative generation process limit their wider application, particularly on resource-constrained platforms like mobile devices. Existing post-training quantization (PTQ) methods have managed to compress diffusion models to low precision. Nevertheless, due to the iterative nature of diffusion models, quantization errors tend to accumulate throughout the generation process. This accumulation of error becomes particularly problematic in low-precision scenarios, leading to significant distortions in the generated images. We attribute this accumulation issue to two main causes: error propagation and exposure bias. To address these problems, we propose a timestep-aware correction method for quantized diffusion model, which dynamically corrects the quantization error. By leveraging the proposed method in low-precision diffusion models, substantial enhancement of output quality could be achieved with only negligible computation overhead. Extensive experiments underscore our method's effectiveness and generalizability. By employing the proposed correction strategy, we achieve state-of-the-art (SOTA) results on low-precision models., Comment: ECCV 2024

Despite their great success across various multimodal tasks, Large Vision-Language Models (LVLMs) are facing a prevalent problem with object hallucinations, where the generated textual responses are inconsistent with ground-truth objects in the given image. This paper investigates various LVLMs and pinpoints attention deficiency toward discriminative local image features as one root cause of object hallucinations. Specifically, LVLMs predominantly attend to prompt-independent global image features, while failing to capture prompt-relevant local features, consequently undermining the visual grounding capacity of LVLMs and leading to hallucinations. To this end, we propose Assembly of Global and Local Attention (AGLA), a training-free and plug-and-play approach that mitigates object hallucinations by exploring an ensemble of global features for response generation and local features for visual discrimination simultaneously. Our approach exhibits an image-prompt matching scheme that captures prompt-relevant local features from images, leading to an augmented view of the input image where prompt-relevant content is reserved while irrelevant distractions are masked. With the augmented view, a calibrated decoding distribution can be derived by integrating generative global features from the original image and discriminative local features from the augmented image. Extensive experiments show that AGLA consistently mitigates object hallucinations and enhances general perception capability for LVLMs across various discriminative and generative benchmarks. Our code will be released at https://github.com/Lackel/AGLA.

Text-to-image diffusion models benefit artists with high-quality image generation. Yet their stochastic nature hinders artists from creating consistent images of the same subject. Existing methods try to tackle this challenge and generate consistent content in various ways. However, they either depend on external restricted data or require expensive tuning of the diffusion model. For this issue, we propose a novel one-shot tuning paradigm, termed OneActor. It efficiently performs consistent subject generation solely driven by prompts via a learned semantic guidance to bypass the laborious backbone tuning. We lead the way to formalize the objective of consistent subject generation from a clustering perspective, and thus design a cluster-conditioned model. To mitigate the overfitting challenge shared by one-shot tuning pipelines, we augment the tuning with auxiliary samples and devise two inference strategies: semantic interpolation and cluster guidance. These techniques are later verified to significantly improve the generation quality. Comprehensive experiments show that our method outperforms a variety of baselines with satisfactory subject consistency, superior prompt conformity as well as high image quality. Our method is capable of multi-subject generation and compatible with popular diffusion extensions. Besides, we achieve a 4 times faster tuning speed than tuning-based baselines and, if desired, avoid increasing the inference time. Furthermore, our method can be naturally utilized to pre-train a consistent subject generation network from scratch, which will implement this research task into more practical applications. (Project page: https://johnneywang.github.io/OneActor-webpage/)

While large-scale pre-trained text-to-image models can synthesize diverse and high-quality human-centered images, novel challenges arise with a nuanced task of "identity fine editing": precisely modifying specific features of a subject while maintaining its inherent identity and context. Existing personalization methods either require time-consuming optimization or learning additional encoders, adept in "identity re-contextualization". However, they often struggle with detailed and sensitive tasks like human face editing. To address these challenges, we introduce DreamSalon, a noise-guided, staged-editing framework, uniquely focusing on detailed image manipulations and identity-context preservation. By discerning editing and boosting stages via the frequency and gradient of predicted noises, DreamSalon first performs detailed manipulations on specific features in the editing stage, guided by high-frequency information, and then employs stochastic denoising in the boosting stage to improve image quality. For more precise editing, DreamSalon semantically mixes source and target textual prompts, guided by differences in their embedding covariances, to direct the model's focus on specific manipulation areas. Our experiments demonstrate DreamSalon's ability to efficiently and faithfully edit fine details on human faces, outperforming existing methods both qualitatively and quantitatively.

Generalized Category Discovery (GCD) is a crucial task that aims to recognize both known and novel categories from a set of unlabeled data by utilizing a few labeled data with only known categories. Due to the lack of supervision and category information, current methods usually perform poorly on novel categories and struggle to reveal semantic meanings of the discovered clusters, which limits their applications in the real world. To mitigate the above issues, we propose Loop, an end-to-end active-learning framework that introduces Large Language Models (LLMs) into the training loop, which can boost model performance and generate category names without relying on any human efforts. Specifically, we first propose Local Inconsistent Sampling (LIS) to select samples that have a higher probability of falling to wrong clusters, based on neighborhood prediction consistency and entropy of cluster assignment probabilities. Then we propose a Scalable Query strategy to allow LLMs to choose true neighbors of the selected samples from multiple candidate samples. Based on the feedback from LLMs, we perform Refined Neighborhood Contrastive Learning (RNCL) to pull samples and their neighbors closer to learn clustering-friendly representations. Finally, we select representative samples from clusters corresponding to novel categories to allow LLMs to generate category names for them. Extensive experiments on three benchmark datasets show that Loop outperforms SOTA models by a large margin and generates accurate category names for the discovered clusters. Code and data are available at https://github.com/Lackel/LOOP., Comment: Accepted by ACL 2024 Findings, code and data are available at https://github.com/Lackel/LOOP

Histological examination is crucial for cancer diagnosis, including hematoxylin and eosin (H&E) staining for mapping morphology and immunohistochemistry (IHC) staining for revealing chemical information. Recently developed two-color stimulated Raman histology could bypass the complex tissue processing to mimic H&E-like morphology. Yet, the underlying chemical features are not revealed, compromising the effectiveness of prognostic stratification. Here, we present a high-content stimulated Raman histology (HC-SRH) platform that provides both morphological and chemical information for cancer diagnosis based on un-stained breast tissues. Through spectral unmixing in the C-H vibration window, HC-SRH can map unsaturated lipids, cellular protein, extracellular matrix, saturated lipid, and water in breast tissue. In this way, HC-SRH provides excellent contrast for various tissue components. Considering rapidness is important in clinical trials, we implemented spectral selective sampling to boost the speed of HC-SRH by one order. We also successfully demonstrated the HC-SRH in a clinical-compatible fiber laser-based SRS microscopy. With the widely rapid tuning capability of the advanced fiber laser, a clear chemical contrast of nucleic acid and solid-state ester is shown in the fingerprint result., Comment: 6 figures

Label-free vibrational imaging by stimulated Raman scattering (SRS) provides unprecedented insight into real-time chemical distributions in living systems. Specifically, SRS in the fingerprint region can resolve multiple chemicals in a complex bio-environment using specific and well-separated Raman signatures. Yet, fingerprint SRS imaging with microsecond spectral acquisition has not been achieved due to the small fingerprint Raman cross-sections and the lack of ultrafast acquisition scheme with high spectral resolution and high fidelity. Here, we report a fingerprint spectroscopic SRS platform that acquires a distortion-free SRS spectrum with 10 cm-1 spectral resolution in 20 microseconds using a lab-built ultrafast delay-line tuning system. Meanwhile, we significantly improve the signal-to-noise ratio by employing a spatial-spectral residual learning network, reaching comparable quality to images taken with two orders of magnitude longer pixel dwell times. Collectively, our system achieves reliable fingerprint spectroscopic SRS with microsecond spectral acquisition speed, enabling imaging and tracking of multiple biomolecules in samples ranging from a live single microbe to a tissue slice, which was not previously possible with SRS imaging in the highly congested carbon-hydrogen region. To show the broad utility of the approach, we have demonstrated high-speed compositional imaging of lipid metabolism in living pancreatic cancer Mia PaCa-2 cells. We then performed high-resolution mapping of cholesterol, fatty acid, and protein in the mouse whole brain. Finally, we mapped the production of two biofuels in microbial samples by harnessing the superior spectral and temporal resolutions of our system., Comment: The manuscript contains 24 pages and 4 main figures. The supplementary information contains 7 figures. Code and data are available at https://github.com/buchenglab

Stimulated Raman scattering (SRS) microscopy allows for high-speed label-free chemical imaging of biomedical systems. The imaging sensitivity of SRS microscopy is limited to ~10 mM for endogenous biomolecules. Electronic pre-resonant SRS allows detection of sub-micromolar chromophores. However, label-free SRS detection of single biomolecules having extremely small Raman cross-sections (~10-30 cm2 sr-1) remains unreachable. Here, we demonstrate plasmon-enhanced stimulated Raman scattering (PESRS) microscopy with single-molecule detection sensitivity. Incorporating pico-Joule laser excitation, background subtraction, and a denoising algorithm, we obtained robust single-pixel SRS spectra exhibiting the statistics of single-molecule events. Single-molecule detection was verified by using two isotopologues of adenine. We further demonstrated the capability of applying PESRS for biological applications and utilized PESRS to map adenine released from bacteria due to starvation stress. PESRS microscopy holds the promise for ultrasensitive detection of molecular events in chemical and biomedical systems.