Recent advances in large pretrained text-to-image models have shown unprecedented capabilities for high-quality human-centric generation, however, customizing face identity is still an intractable problem. Existing methods cannot ensure stable identity preservation and flexible editability, even with several images for each subject during training. In this work, we propose StableIdentity, which allows identity-consistent recontextualization with just one face image. More specifically, we employ a face encoder with an identity prior to encode the input face, and then land the face representation into a space with an editable prior, which is constructed from celeb names. By incorporating identity prior and editability prior, the learned identity can be injected anywhere with various contexts. In addition, we design a masked two-phase diffusion loss to boost the pixel-level perception of the input face and maintain the diversity of generation. Extensive experiments demonstrate our method outperforms previous customization methods. In addition, the learned identity can be flexibly combined with the off-the-shelf modules such as ControlNet. Notably, to the best knowledge, we are the first to directly inject the identity learned from a single image into video/3D generation without finetuning. We believe that the proposed StableIdentity is an important step to unify image, video, and 3D customized generation models.
Imaging Atmospheric Cherenkov Telescopes (IACTs) of gamma ray observatory TAIGA detect the Extesnive Air Showers (EASs) originating from the cosmic or gamma rays interactions with the atmosphere. Thereby, telescopes obtain images of the EASs. The ability to segregate gamma rays images from the hadronic cosmic ray background is one of the main features of this type of detectors. However, in actual IACT observations simultaneous observation of the background and the source of gamma ray is needed. This observation mode (called wobbling) modifies images of events, which affects the quality of selection by neural networks. Thus, in this work, the results of the application of neural networks (NN) for image classification task on Monte Carlo (MC) images of TAIGA-IACTs are presented. The wobbling mode is considered together with the image adaptation for adequate analysis by NNs. Simultaneously, we explore several neural network structures that classify events both directly from images or through Hillas parameters extracted from images. In addition, by employing NNs, MC simulation data are used to evaluate the quality of the segregation of rare gamma events with the account of all necessary image modifications.
3D human pose estimation captures the human joint points in three-dimensional space while keeping the depth information and physical structure. That is essential for applications that require precise pose information, such as human-computer interaction, scene understanding, and rehabilitation training. Due to the challenges in data collection, mainstream datasets of 3D human pose estimation are primarily composed of multi-view video data collected in laboratory environments, which contains rich spatial-temporal correlation information besides the image frame content. Given the remarkable self-attention mechanism of transformers, capable of capturing the spatial-temporal correlation from multi-view video datasets, we propose a multi-stage framework for 3D sequence-to-sequence (seq2seq) human pose detection. Firstly, the spatial module represents the human pose feature by intra-image content, while the frame-image relation module extracts temporal relationships and 3D spatial positional relationship features between the multi-perspective images. Secondly, the self-attention mechanism is adopted to eliminate the interference from non-human body parts and reduce computing resources. Our method is evaluated on Human3.6M, a popular 3D human pose detection dataset. Experimental results demonstrate that our approach achieves state-of-the-art performance on this dataset.
While traditional video representations are organized around discrete image frames, event-based video is a new paradigm that forgoes image frames altogether. Rather, pixel samples are temporally asynchronous and independent of one another. Until now, researchers have lacked a cohesive software framework for exploring the representation, compression, and applications of event-based video. I present the AD$\Delta$ER software suite to fill this gap. This framework includes utilities for transcoding framed and multimodal event-based video sources to a common representation, rate control mechanisms, lossy compression, application support, and an interactive GUI for transcoding and playback. In this paper, I describe these various software components and their usage.
Recently, significant progress has been made on Large Vision-Language Models (LVLMs); a new class of VL models that make use of large pre-trained language models. Yet, their vulnerability to Typographic attacks, which involve superimposing misleading text onto an image remain unstudied. Furthermore, prior work typographic attacks rely on sampling a random misleading class from a predefined set of classes. However, the random chosen class might not be the most effective attack. To address these issues, we first introduce a novel benchmark uniquely designed to test LVLMs vulnerability to typographic attacks. Furthermore, we introduce a new and more effective typographic attack: Self-Generated typographic attacks. Indeed, our method, given an image, make use of the strong language capabilities of models like GPT-4V by simply prompting them to recommend a typographic attack. Using our novel benchmark, we uncover that typographic attacks represent a significant threat against LVLM(s). Furthermore, we uncover that typographic attacks recommended by GPT-4V using our new method are not only more effective against GPT-4V itself compared to prior work attacks, but also against a host of less capable yet popular open source models like LLaVA, InstructBLIP, and MiniGPT4.
Tensor network structure search (TN-SS), aiming at searching for suitable tensor network (TN) structures in representing high-dimensional problems, largely promotes the efficacy of TN in various machine learning applications. Nonetheless, finding a satisfactory TN structure using existing algorithms remains challenging. To develop more effective algorithms and avoid the human labor-intensive development process, we explore the knowledge embedded in large language models (LLMs) for the automatic design of TN-SS algorithms. Our approach, dubbed GPTN-SS, leverages an elaborate crafting LLM-based prompting system that operates in an evolutionary-like manner. The experimental results, derived from real-world data, demonstrate that GPTN-SS can effectively leverage the insights gained from existing methods to develop novel TN-SS algorithms that achieve a better balance between exploration and exploitation. These algorithms exhibit superior performance in searching the high-quality TN structures for natural image compression and model parameters compression while also demonstrating generalizability in their performance.
In this paper, we present a fully automatic brain tumor segmentation and classification model using a Deep Convolutional Neural Network that includes a multiscale approach. One of the differences of our proposal with respect to previous works is that input images are processed in three spatial scales along different processing pathways. This mechanism is inspired in the inherent operation of the Human Visual System. The proposed neural model can analyze MRI images containing three types of tumors: meningioma, glioma, and pituitary tumor, over sagittal, coronal, and axial views and does not need preprocessing of input images to remove skull or vertebral column parts in advance. The performance of our method on a publicly available MRI image dataset of 3064 slices from 233 patients is compared with previously classical machine learning and deep learning published methods. In the comparison, our method remarkably obtained a tumor classification accuracy of 0.973, higher than the other approaches using the same database.
Low-light image enhancement techniques have significantly progressed, but unstable image quality recovery and unsatisfactory visual perception are still significant challenges. To solve these problems, we propose a novel and robust low-light image enhancement method via CLIP-Fourier Guided Wavelet Diffusion, abbreviated as CFWD. Specifically, we design a guided network with a multiscale visual language in the frequency domain based on the wavelet transform to achieve effective image enhancement iteratively. In addition, we combine the advantages of Fourier transform in detail perception to construct a hybrid frequency domain space with significant perceptual capabilities(HFDPM). This operation guides wavelet diffusion to recover the fine-grained structure of the image and avoid diversity confusion. Extensive quantitative and qualitative experiments on publicly available real-world benchmarks show that our method outperforms existing state-of-the-art methods and better reproduces images similar to normal images. Code is available at https://github.com/He-Jinhong/CFWD.
Despite the remarkable capabilities of deep neural networks in image recognition, the dependence on activation functions remains a largely unexplored area and has yet to be eliminated. On the other hand, Polynomial Networks is a class of models that does not require activation functions, but have yet to perform on par with modern architectures. In this work, we aim close this gap and propose MONet, which relies solely on multilinear operators. The core layer of MONet, called Mu-Layer, captures multiplicative interactions of the elements of the input token. MONet captures high-degree interactions of the input elements and we demonstrate the efficacy of our approach on a series of image recognition and scientific computing benchmarks. The proposed model outperforms prior polynomial networks and performs on par with modern architectures. We believe that MONet can inspire further research on models that use entirely multilinear operations.
Recent works demonstrate that using reinforcement learning (RL) with quality rewards can enhance the quality of generated images in text-to-image (T2I) generation. However, a simple aggregation of multiple rewards may cause over-optimization in certain metrics and degradation in others, and it is challenging to manually find the optimal weights. An effective strategy to jointly optimize multiple rewards in RL for T2I generation is highly desirable. This paper introduces Parrot, a novel multi-reward RL framework for T2I generation. Through the use of the batch-wise Pareto optimal selection, Parrot automatically identifies the optimal trade-off among different rewards during the RL optimization of the T2I generation. Additionally, Parrot employs a joint optimization approach for the T2I model and the prompt expansion network, facilitating the generation of quality-aware text prompts, thus further enhancing the final image quality. To counteract the potential catastrophic forgetting of the original user prompt due to prompt expansion, we introduce original prompt centered guidance at inference time, ensuring that the generated image remains faithful to the user input. Extensive experiments and a user study demonstrate that Parrot outperforms several baseline methods across various quality criteria, including aesthetics, human preference, image sentiment, and text-image alignment.