Leveraging the large generative prior of the flow transformer for tuning-free image editing requires authentic inversion to project the image into the model’s domain and a flexible invariance control mechanism to preserve non-target contents. However, the prevailing diffusion inversion performs deficiently in flow-based models, and the invariance control cannot reconcile diverse rigid and non-rigid editing tasks. To address these, we systematically analyze the inversion and invariance control based on the flow transformer. Specifically, we unveil that the Euler inversion shares a similar structure to DDIM yet is more susceptible to the approximation error. Thus, we propose a two-stage inversion to first refine the velocity estimation and then compensate for the leftover error, which pivots closely to the model prior and benefits editing. Meanwhile, we propose the invariance control that manipulates the text features within the adaptive layer normalization, connecting the changes in the text prompt to image semantics. This mechanism can simultaneously preserve the non-target contents while allowing rigid and non-rigid manipulation, enabling a wide range of editing types such as visual text, quantity, facial expression, etc. Experiments on versatile scenarios validate that our framework achieves flexible and accurate editing, unlocking the potential of the flow transformer for versatile image editing. Project Page is here.

Although diffusion-based image virtual try-on has made considerable progress, emerging approaches still struggle to effectively address the issue of hand occlusion (i.e., clothing regions occluded by the hand part), leading to a notable degradation of the try-on performance. To tackle this issue widely existing in real-world scenarios, we propose VTON-HandFit, leveraging the power of hand priors to reconstruct the appearance and structure for hand occlusion cases. Firstly, we tailor a Hand-Pose Aggregation Net using the ControlNet-based structure explicitly and adaptively encoding the global hand and pose priors. Besides, to fully exploit the hand-related structure and appearance information, we propose Hand-feature Disentanglement Embedding module to disentangle the hand priors into the hand structure-parametric and visual-appearance features, and customize a masked cross attention for further decoupled feature embedding. Lastly, we customize a hand-canny constraint loss to better learn the structure edge knowledge from the hand template of model image. VTON-HandFit outperforms the baselines in qualitative and quantitative evaluations on the public dataset and our self-collected hand-occlusion Handfit-3K dataset particularly for the arbitrary hand pose occlusion cases in real-world scenarios. Our project page is at: https://vton-handfit.github.io.

With the continuous progress of visual generation technologies, the scale of video datasets has grown exponentially. The quality of these datasets plays a pivotal role in the performance of video generation models. We assert that temporal splitting, detailed captions, and video quality filtering are three crucial determinants of dataset quality. However, existing datasets exhibit various limitations in these areas. To address these challenges, we introduce Koala-36M, a large-scale, high-quality video dataset featuring accurate temporal splitting, detailed captions, and superior video quality. The essence of our approach lies in improving the consistency between fine-grained conditions and video content. Specifically, we employ a linear classifier on probability distributions to enhance the accuracy of transition detection, ensuring better temporal consistency. We then provide structured captions for the splitted videos, with an average length of 200 words, to improve text-video alignment. Additionally, we develop a Video Training Suitability Score (VTSS) that integrates multiple sub-metrics, allowing us to filter high-quality videos from the original corpus. Finally, we incorporate several metrics into the training process of the generation model, further refining the fine-grained conditions. Our experiments demonstrate the effectiveness of our data processing pipeline and the quality of the proposed Koala-36M dataset. Our dataset and code have been released at https://koala36m.github.io/.

Despite remarkable progress in image-based virtual try-on systems, generating realistic and robust fitting images for cross-category virtual try-on remains a challenging task. The primary difficulty arises from the absence of human-like reasoning, which involves addressing size mismatches between garments and models while recognizing and leveraging the distinct functionalities of various regions within the model images. To address this issue, we draw inspiration from human cognitive processes and disentangle the complex reasoning required for cross-category try-on into a structured framework. This framework systematically decomposes the model image into three distinct regions: try-on, reconstruction, and imagination zones. Each zone plays a specific role in accommodating the garment and facilitating realistic synthesis. To endow the model with robust reasoning capabilities for cross-category scenarios, we propose an iterative data constructor. This constructor encompasses diverse scenarios, including intra-category try-on, any-to-dress transformations (replacing any garment category with a dress), and dress-to-any transformations (replacing a dress with another garment category). Utilizing the generated dataset, we introduce a tri-zone priors generator that intelligently predicts the try-on, reconstruction, and imagination zones by analyzing how the input garment is expected to align with the model image. Guided by these tri-zone priors, our proposed method, CrossVTON, achieves state-of-the-art performance, surpassing existing baselines in both qualitative and quantitative evaluations. Notably, it demonstrates superior capability in handling cross-category virtual try-on, meeting the complex demands of real-world applications.

Modern video generation frameworks based on Latent Diffusion Models suffer from inefficiencies in tokenization due to the Frame-Proportional Information Assumption. Existing tokenizers provide fixed temporal compression rates, causing the computational cost of the diffusion model to scale linearly with the frame rate. The paper proposes the Duration-Proportional Information Assumption: the upper bound on the information capacity of a video is proportional to the duration rather than the number of frames. Based on this insight, the paper introduces VFRTok, a Transformer-based video tokenizer, that enables variable frame rate encoding and decoding through asymmetric frame rate training between the encoder and decoder. Furthermore, the paper proposes Partial Rotary Position Embeddings (RoPE) to decouple position and content modeling, which groups correlated patches into unified tokens. The Partial RoPE effectively improves content-awareness, enhancing the video generation capability. Benefiting from the compact and continuous spatio-temporal representation, VFRTok achieves competitive reconstruction quality and state-of-the-art generation fidelity while using only $1/8$ tokens compared to existing tokenizers.

Recently, the rise of large-scale vision-language pretrained models like CLIP, coupled with the technology of Parameter-Efficient FineTuning (PEFT), has captured substantial attraction in video action recognition. Nevertheless, prevailing approaches tend to prioritize strong supervised performance at the expense of compromising the models' generalization capabilities during transfer. In this paper, we introduce a novel Multimodal, Multi-task CLIP adapting framework named M2-CLIP to address these challenges, preserving both high supervised performance and robust transferability. Firstly, to enhance the individual modality architectures, we introduce multimodal adapters to both the visual and text branches. Specifically, we design a novel visual TED-Adapter, that performs global Temporal Enhancement and local temporal Difference modeling to improve the temporal representation capabilities of the visual encoder. Moreover, we adopt text encoder adapters to strengthen the learning of semantic label information. Secondly, we design a multi-task decoder with a rich set of supervisory signals, including the original contrastive learning head, a cross-modal classification head, a cross-modal masked language modeling head, and a visual classification head. This multi-task decoder adeptly satisfies the need for strong supervised performance within a multimodal framework. Experimental results validate the efficacy of our approach, demonstrating exceptional performance in supervised learning while maintaining strong generalization in zero-shot scenarios.

Recent advancements in personalized image generation using diffusion models have been noteworthy. However, existing methods suffer from inefficiencies due to the requirement for subject-specific fine-tuning. This computationally intensive process hinders efficient deployment, limiting practical usability. Moreover, these methods often grapple with identity distortion and limited expression diversity. In light of these challenges, we propose Portrait-Booth, an innovative approach designed for high efficiency, robust identity preservation, and expression-editable text-to-image generation, without the need for fine-tuning. Por-traitBooth leverages subject embeddingsfrom aface recognition model for personalized image generation without fine-tuning. It eliminates computational overhead and mitigates identity distortion. The introduced dynamic identity preservation strategy further ensures close resemblance to the original image identity. Moreover, PortraitBooth incorporates emotion-aware cross-attention control for diverse facial expressions in generated images, supporting text-driven expression editing. Its scalability enables efficient and high-quality image creation, including multi-subject generation. Extensive results demonstrate superior performance over other state-of-the-art methods in both single and multiple image generation scenarios. Our project page is at https://portraitbooth.github.io.

Motion retargeting is a fundamental problem in computer graphics and computer vision. Existing approaches usually have many strict requirements, such as the source-target skeletons needing to have the same number of joints or share the same topology. To tackle this problem, we note that skeletons with different structure may have some common body parts despite the differences in joint numbers. Following this observation, we propose a novel, flexible motion retargeting framework. The key idea of our method is to regard the body part as the basic retargeting unit rather than directly retargeting the whole body motion. To enhance the spatial modeling capability of the motion encoder, we introduce a pose-aware attention network (PAN) in the motion encoding phase. The PAN is pose-aware since it can dynamically predict the joint weights within each body part based on the input pose, and then construct a shared latent space for each body part by feature pooling. Extensive experiments show that our approach can generate better motion retargeting results both qualitatively and quantitatively than state-of-the-art methods. Moreover, we also show that our framework can generate reasonable results even for a more challenging retargeting scenario, like retargeting between bipedal and quadrupedal skeletons because of the body part retargeting strategy and PAN.

3D human pose estimation has been a long-standing challenge in computer vision and graphics, where multi-view methods have significantly progressed but are limited by the tedious calibration processes. Existing multi-view methods are restricted to fixed camera pose and therefore lack generalization ability. This paper presents a novel Probabilistic Triangulation module that can be embedded in a calibrated 3D human pose estimation method, generalizing it to uncalibration scenes. The key idea is to use a probability distribution to model the camera pose and iteratively update the distribution from 2D features instead of using camera pose. Specifically, We maintain a camera pose distribution and then iteratively update this distribution by computing the posterior probability of the camera pose through Monte Carlo sampling. This way, the gradients can be directly back-propagated from the 3D pose estimation to the 2D heatmap, enabling end-to-end training. Extensive experiments on Human3.6M and CMU Panoptic demonstrate that our method outperforms other uncalibration methods and achieves comparable results with state-of-the-art calibration methods. Thus, our method achieves a trade-off between estimation accuracy and generalizability. Our code is in https://github.com/bymaths/probabilistictriangulation

Video frame interpolation is an important low-level vision task, which can increase frame rate for more fluent visual experience. Existing methods have achieved great success by employing advanced motion models and synthesis networks. However, the spatial redundancy when synthesizing the target frame has not been fully explored, that can result in lots of inefficient computation. On the other hand, the computation compression degree in frame interpolation is highly dependent on both texture distribution and scene motion, which demands to understand the spatial-temporal information of each input frame pair for a better compression degree selection. In this work, we propose a novel two-stage frame interpolation framework termed WaveletVFI to address above problems. It first estimates intermediate optical flow with a lightweight motion perception network, and then a wavelet synthesis network uses flow aligned context features to predict multi-scale wavelet coefficients with sparse convolution for efficient target frame reconstruction, where the sparse valid masks that control computation in each scale are determined by a crucial threshold ratio. Instead of setting a fixed value like previous methods, we find that embedding a classifier in the motion perception network to learn a dynamic threshold for each sample can achieve more computation reduction with almost no loss of accuracy. On the common high resolution and animation frame interpolation benchmarks, proposed WaveletVFI can reduce computation up to 40% while maintaining similar accuracy, making it perform more efficiently against other state-of-the-arts.

Prevailing video frame interpolation algorithms, that generate the intermediate frames from consecutive inputs, typically rely on complex model architectures with heavy parameters or large delay, hindering them from diverse real-time applications. In this work, we devise an efficient encoder-decoder based network, termed IFRNet, for fast in-termediate frame synthesizing. It first extracts pyramid features from given inputs, and then refines the bilateral in-termediate flow fields together with a powerful intermedi-ate feature until generating the desired output. The gradu-ally refined intermediate feature can not only facilitate in-termediate flow estimation, but also compensate for con-textual details, making IFRNet do not need additional syn-thesis or refinement module. To fully release its potential, we further propose a novel task-oriented optical flow dis-tillation loss to focus on learning the useful teacher knowl-edge towards frame synthesizing. Meanwhile, a new ge-ometry consistency regularization term is imposed on the gradually refined intermediate features to keep better structure layout. Experiments on various benchmarks demon-strate the excellent performance and fast inference speed of proposed approaches. Code is available at https://github.com/ltkong218/IFRNet.

For action recognition learning, 2D CNN-based methods are efficient but may yield redundant features due to applying the same 2D convolution kernel to each frame. Recent efforts attempt to capture motion information by establishing inter-frame connections while still suffering the limited temporal receptive field or high latency. Moreover, the feature enhancement is often only performed by channel or space dimension in action recognition. To address these issues, we first devise a Channel-wise Motion Enhancement (CME) module to adaptively emphasize the channels related to dynamic information with a channel-wise gate vector. The channel gates generated by CME incorporate the information from all the other frames in the video. We further propose a Spatial-wise Motion Enhancement (SME) module to focus on the regions with the critical target in motion, according to the point-to-point similarity between adjacent feature maps. The intuition is that the change of background is typically slower than the motion area. Both CME and SME have clear physical meaning in capturing action clues. By integrating the two modules into the off-the-shelf 2D network, we finally obtain a Comprehensive Motion Representation (CMR) learning method for action recognition, which achieves competitive performance on Something-Something V1 & V2 and Kinetics-400. On the temporal reasoning datasets Something-Something V1 and V2, our method outperforms the current state-of-the-art by 2.3% and 1.9% when using 16 frames as input, respectively.

In facial action unit (AU) recognition tasks, regional feature learning and AU relation modeling are two effective aspects which are worth exploring. However, the limited representation capacity of regional features makes it difficult for relation models to embed AU relationship knowledge. In this paper, we propose a novel multi-level adaptive ROI and graph learning (MARGL) framework to tackle this problem. Specifically, an adaptive ROI learning module is designed to automatically adjust the location and size of the predefined AU regions. Meanwhile, besides relationship between AUs, there exists strong relevance between regional features across multiple levels of the backbone network as level-wise features focus on different aspects of representation. In order to incorporate the intra-level AU relation and inter-level AU regional relevance simultaneously, a multi-level AU relation graph is constructed and graph convolution is performed to further enhance AU regional features of each level. Experiments on BP4D and DISFA demonstrate the proposed MARGL significantly outperforms the previous state-of-the-art methods.

One crucial aspect of partial domain adaptation (PDA) is how to select the relevant source samples in the shared classes for knowledge transfer. Previous PDA methods tackle this problem by re-weighting the source samples based on their high-level information (deep features). However, since the domain shift between source and target domains, only using the deep features for sample selection is defective. We argue that it is more reasonable to additionally exploit the pixel-level information for PDA problem, as the appearance difference between outlier source classes and target classes is significantly large. In this paper, we propose a reinforced transfer network (RTNet), which utilizes both high-level and pixel-level information for PDA problem. Our RTNet is composed of a reinforced data selector (RDS) based on reinforcement learning (RL), which filters out the outlier source samples, and a domain adaptation model which minimizes the domain discrepancy in the shared label space. Specifically, in the RDS, we design a novel reward based on the reconstruct errors of selected source samples on the target generator, which introduces the pixel-level information to guide the learning of RDS. Besides, we develope a state containing high-level information, which used by the RDS for sample selection. The proposed RDS is a general module, which can be easily integrated into existing DA models to make them fit the PDA situation. Extensive experiments indicate that RTNet can achieve state-of-the-art performance for PDA tasks on several benchmark datasets.

Recently, considerable effort has been devoted to deep domain adaptation in computer vision and machine learning communities. However, most of existing work only concentrates on learning shared feature representation by minimizing the distribution discrepancy across different domains. Due to the fact that all the domain alignment approaches can only reduce, but not remove the domain shift, target domain samples distributed near the edge of the clusters, or far from their corresponding class centers are easily to be misclassified by the hyperplane learned from the source domain. To alleviate this issue, we propose to joint domain alignment and discriminative feature learning, which could benefit both domain alignment and final classification. Specifically, an instance-based discriminative feature learning method and a center-based discriminative feature learning method are proposed, both of which guarantee the domain invariant features with better intra-class compactness and inter-class separability. Extensive experiments show that learning the discriminative features in the shared feature space can significantly boost the performance of deep domain adaptation methods.

Spatiotemporal and motion features are two complementary and crucial information for video action recognition. Recent state-of-the-art methods adopt a 3D CNN stream to learn spatiotemporal features and another flow stream to learn motion features. In this work, we aim to efficiently encode these two features in a unified 2D framework. To this end, we first propose a STM block, which contains a Channel-wise SpatioTemporal Module (CSTM) to present the spatiotemporal features and a Channel-wise Motion Module (CMM) to efficiently encode motion features. We then replace original residual blocks in the ResNet architecture with STM blcoks to form a simple yet effective STM network by introducing very limited extra computation cost. Extensive experiments demonstrate that the proposed STM network outperforms the state-of-the-art methods on both temporal-related datasets (i.e., Something-Something v1 & v2 and Jester) and scene-related datasets (i.e., Kinetics-400, UCF-101, and HMDB-51) with the help of encoding spatiotemporal and motion features together.

Recent years, transfer learning has attracted much attention in the community of machine learning. In this paper, we mainly focus on the tasks of parameter transfer under the framework of extreme learning machine (ELM). Unlike the existing parameter transfer approaches, which incorporate the source model information into the target by regularizing the difference between the source and target domain parameters, an intuitively appealing projective-model is proposed to bridge the source and target model parameters. Specifically, we formulate the parameter transfer in the ELM networks by the means of parameter projection, and train the model by optimizing the projection matrix and classifier parameters jointly. Further more, the ℓ2,1-norm structured sparsity penalty is imposed on the source domain parameters, which encourages the joint feature selection and parameter transfer. To evaluate the effectiveness of the proposed method, comprehensive experiments on several commonly used domain adaptation datasets are presented. The results show that the proposed method significantly outperforms the non-transfer ELM networks and other classical transfer learning methods.