Human-object interaction (HOI) detection aims to locate human-object pairs and identify their interaction categories in images. Most existing methods primarily focus on supervised learning, which relies on extensive manual HOI annotations. In this paper, we propose a novel framework, termed Knowledge Integration to HOI (KI2HOI), that effectively integrates the knowledge of visual-language model to improve zero-shot HOI detection. Specifically, the verb feature learning module is designed based on visual semantics, by employing the verb extraction decoder to convert corresponding verb queries into interaction-specific category representations. We develop an effective additive self-attention mechanism to generate more comprehensive visual representations. Moreover, the innovative interaction representation decoder effectively extracts informative regions by integrating spatial and visual feature information through a cross-attention mechanism. To deal with zero-shot learning in low-data, we leverage a priori knowledge from the CLIP text encoder to initialize the linear classifier for enhanced interaction understanding. Extensive experiments conducted on the mainstream HICO-DET and V-COCO datasets demonstrate that our model outperforms the previous methods in various zero-shot and full-supervised settings.
We propose VLOGGER, a method for audio-driven human video generation from a single input image of a person, which builds on the success of recent generative diffusion models. Our method consists of 1) a stochastic human-to-3d-motion diffusion model, and 2) a novel diffusion-based architecture that augments text-to-image models with both spatial and temporal controls. This supports the generation of high quality video of variable length, easily controllable through high-level representations of human faces and bodies. In contrast to previous work, our method does not require training for each person, does not rely on face detection and cropping, generates the complete image (not just the face or the lips), and considers a broad spectrum of scenarios (e.g. visible torso or diverse subject identities) that are critical to correctly synthesize humans who communicate. We also curate MENTOR, a new and diverse dataset with 3d pose and expression annotations, one order of magnitude larger than previous ones (800,000 identities) and with dynamic gestures, on which we train and ablate our main technical contributions. VLOGGER outperforms state-of-the-art methods in three public benchmarks, considering image quality, identity preservation and temporal consistency while also generating upper-body gestures. We analyze the performance of VLOGGER with respect to multiple diversity metrics, showing that our architectural choices and the use of MENTOR benefit training a fair and unbiased model at scale. Finally we show applications in video editing and personalization.
Learning-based Text-to-Image (TTI) models like Stable Diffusion have revolutionized the way visual content is generated in various domains. However, recent research has shown that nonnegligible social bias exists in current state-of-the-art TTI systems, which raises important concerns. In this work, we target resolving the social bias in TTI diffusion models. We begin by formalizing the problem setting and use the text descriptions of bias groups to establish an unsafe direction for guiding the diffusion process. Next, we simplify the problem into a weight optimization problem and attempt a Reinforcement solver, Policy Gradient, which shows sub-optimal performance with slow convergence. Further, to overcome limitations, we propose an iterative distribution alignment (IDA) method. Despite its simplicity, we show that IDA shows efficiency and fast convergence in resolving the social bias in TTI diffusion models. Our code will be released.
Diffusion models have demonstrated exceptional capability in generating high-quality images, videos, and audio. Due to their adaptiveness in iterative refinement, they provide a strong potential for achieving better non-autoregressive sequence generation. However, existing text diffusion models still fall short in their performance due to a challenge in handling the discreteness of language. This paper thoroughly analyzes text diffusion models and uncovers two significant limitations: degradation of self-conditioning during training and misalignment between training and sampling. Motivated by our findings, we propose a novel Text Diffusion model called TREC, which mitigates the degradation with Reinforced Conditioning and the misalignment by Time-Aware Variance Scaling. Our extensive experiments demonstrate the competitiveness of TREC against autoregressive, non-autoregressive, and diffusion baselines. Moreover, qualitative analysis shows its advanced ability to fully utilize the diffusion process in refining samples.
Intent is defined for understanding spoken language in existing works. Both textual features and acoustic features involved in medical speech contain intent, which is important for symptomatic diagnosis. In this paper, we propose a medical speech classification model named DRSC that automatically learns to disentangle intent and content representations from textual-acoustic data for classification. The intent representations of the text domain and the Mel-spectrogram domain are extracted via intent encoders, and then the reconstructed text feature and the Mel-spectrogram feature are obtained through two exchanges. After combining the intent from two domains into a joint representation, the integrated intent representation is fed into a decision layer for classification. Experimental results show that our model obtains an average accuracy rate of 95% in detecting 25 different medical symptoms.
The performance of CLIP in dynamic facial expression recognition (DFER) task doesn't yield exceptional results as observed in other CLIP-based classification tasks. While CLIP's primary objective is to achieve alignment between images and text in the feature space, DFER poses challenges due to the abstract nature of text and the dynamic nature of video, making label representation limited and perfect alignment difficult. To address this issue, we have designed A$^{3}$lign-DFER, which introduces a new DFER labeling paradigm to comprehensively achieve alignment, thus enhancing CLIP's suitability for the DFER task. Specifically, our A$^{3}$lign-DFER method is designed with multiple modules that work together to obtain the most suitable expanded-dimensional embeddings for classification and to achieve alignment in three key aspects: affective, dynamic, and bidirectional. We replace the input label text with a learnable Multi-Dimensional Alignment Token (MAT), enabling alignment of text to facial expression video samples in both affective and dynamic dimensions. After CLIP feature extraction, we introduce the Joint Dynamic Alignment Synchronizer (JAS), further facilitating synchronization and alignment in the temporal dimension. Additionally, we implement a Bidirectional Alignment Training Paradigm (BAP) to ensure gradual and steady training of parameters for both modalities. Our insightful and concise A$^{3}$lign-DFER method achieves state-of-the-art results on multiple DFER datasets, including DFEW, FERV39k, and MAFW. Extensive ablation experiments and visualization studies demonstrate the effectiveness of A$^{3}$lign-DFER. The code will be available in the future.
Existing generative adversarial network (GAN) based conditional image generative models typically produce fixed output for the same conditional input, which is unreasonable for highly subjective tasks, such as large-mask image inpainting or style transfer. On the other hand, GAN-based diverse image generative methods require retraining/fine-tuning the network or designing complex noise injection functions, which is computationally expensive, task-specific, or struggle to generate high-quality results. Given that many deterministic conditional image generative models have been able to produce high-quality yet fixed results, we raise an intriguing question: is it possible for pre-trained deterministic conditional image generative models to generate diverse results without changing network structures or parameters? To answer this question, we re-examine the conditional image generation tasks from the perspective of adversarial attack and propose a simple and efficient plug-in projected gradient descent (PGD) like method for diverse and controllable image generation. The key idea is attacking the pre-trained deterministic generative models by adding a micro perturbation to the input condition. In this way, diverse results can be generated without any adjustment of network structures or fine-tuning of the pre-trained models. In addition, we can also control the diverse results to be generated by specifying the attack direction according to a reference text or image. Our work opens the door to applying adversarial attack to low-level vision tasks, and experiments on various conditional image generation tasks demonstrate the effectiveness and superiority of the proposed method.
Text-to-image diffusion models have achieved remarkable performance in image synthesis, while the text interface does not always provide fine-grained control over certain image factors. For instance, changing a single token in the text can have unintended effects on the image. This paper shows a simple modification of classifier-free guidance can help disentangle image factors in text-to-image models. The key idea of our method, Contrastive Guidance, is to characterize an intended factor with two prompts that differ in minimal tokens: the positive prompt describes the image to be synthesized, and the baseline prompt serves as a "baseline" that disentangles other factors. Contrastive Guidance is a general method we illustrate whose benefits in three scenarios: (1) to guide domain-specific diffusion models trained on an object class, (2) to gain continuous, rig-like controls for text-to-image generation, and (3) to improve the performance of zero-shot image editors.
Current makeup transfer methods are limited to simple makeup styles, making them difficult to apply in real-world scenarios. In this paper, we introduce Stable-Makeup, a novel diffusion-based makeup transfer method capable of robustly transferring a wide range of real-world makeup, onto user-provided faces. Stable-Makeup is based on a pre-trained diffusion model and utilizes a Detail-Preserving (D-P) makeup encoder to encode makeup details. It also employs content and structural control modules to preserve the content and structural information of the source image. With the aid of our newly added makeup cross-attention layers in U-Net, we can accurately transfer the detailed makeup to the corresponding position in the source image. After content-structure decoupling training, Stable-Makeup can maintain content and the facial structure of the source image. Moreover, our method has demonstrated strong robustness and generalizability, making it applicable to varioustasks such as cross-domain makeup transfer, makeup-guided text-to-image generation and so on. Extensive experiments have demonstrated that our approach delivers state-of-the-art (SOTA) results among existing makeup transfer methods and exhibits a highly promising with broad potential applications in various related fields.
As more than 70$\%$ of reviews in the existing opinion summary data set are positive, current opinion summarization approaches are reluctant to generate negative summaries given the input of negative texts. To address such sentiment bias, a direct approach without the over-reliance on a specific framework is to generate additional data based on large language models to balance the emotional distribution of the dataset. However, data augmentation based on large language models faces two disadvantages: 1) the potential issues or toxicity in the augmented data; 2) the expensive costs. Therefore, in this paper, we propose a novel data augmentation framework based on both large and small language models for debiasing opinion summarization. In specific, a small size of synthesized negative reviews is obtained by rewriting the positive text via a large language model. Then, a disentangle reconstruction model is trained based on the generated data. After training, a large amount of synthetic data can be obtained by decoding the new representation obtained from the combination of different sample representations and filtering based on confusion degree and sentiment classification. Experiments have proved that our framework can effectively alleviate emotional bias same as using only large models, but more economically.