AI-generated content (AIGC) methods aim to produce text, images, videos, 3D assets, and other media using AI algorithms. Due to its wide range of applications and the demonstrated potential of recent works, AIGC developments have been attracting a lot of attention recently, and AIGC methods have been developed for various data modalities, such as image, video, text, 3D shape (as voxels, point clouds, meshes, and neural implicit fields), 3D scene, 3D human avatar (body and head), 3D motion, and audio -- each presenting different characteristics and challenges. Furthermore, there have also been many significant developments in cross-modality AIGC methods, where generative methods can receive conditioning input in one modality and produce outputs in another. Examples include going from various modalities to image, video, 3D shape, 3D scene, 3D avatar (body and head), 3D motion (skeleton and avatar), and audio modalities. In this paper, we provide a comprehensive review of AIGC methods across different data modalities, including both single-modal and cross-modality methods, highlighting the various challenges, representative works, and recent technical directions in each setting. We also present comparative results on several benchmark datasets in various modalities. Moreover, we also discuss the challenges and potential future research directions.
Reconstructing 3D objects from extremely sparse views is a long-standing and challenging problem. While recent techniques employ image diffusion models for generating plausible images at novel viewpoints or for distilling pre-trained diffusion priors into 3D representations using score distillation sampling (SDS), these methods often struggle to simultaneously achieve high-quality, consistent, and detailed results for both novel-view synthesis (NVS) and geometry. In this work, we present Sparse3D, a novel 3D reconstruction method tailored for sparse view inputs. Our approach distills robust priors from a multiview-consistent diffusion model to refine a neural radiance field. Specifically, we employ a controller that harnesses epipolar features from input views, guiding a pre-trained diffusion model, such as Stable Diffusion, to produce novel-view images that maintain 3D consistency with the input. By tapping into 2D priors from powerful image diffusion models, our integrated model consistently delivers high-quality results, even when faced with open-world objects. To address the blurriness introduced by conventional SDS, we introduce the category-score distillation sampling (C-SDS) to enhance detail. We conduct experiments on CO3DV2 which is a multi-view dataset of real-world objects. Both quantitative and qualitative evaluations demonstrate that our approach outperforms previous state-of-the-art works on the metrics regarding NVS and geometry reconstruction.
The absolute depth values of surrounding environments provide crucial cues for various assistive technologies, such as localization, navigation, and 3D structure estimation. We propose that accurate depth estimated from panoramic images can serve as a powerful and light-weight input for a wide range of downstream tasks requiring 3D information. While panoramic images can easily capture the surrounding context from commodity devices, the estimated depth shares the limitations of conventional image-based depth estimation; the performance deteriorates under large domain shifts and the absolute values are still ambiguous to infer from 2D observations. By taking advantage of the holistic view, we mitigate such effects in a self-supervised way and fine-tune the network with geometric consistency during the test phase. Specifically, we construct a 3D point cloud from the current depth prediction and project the point cloud at various viewpoints or apply stretches on the current input image to generate synthetic panoramas. Then we minimize the discrepancy of the 3D structure estimated from synthetic images without collecting additional data. We empirically evaluate our method in robot navigation and map-free localization where our method shows large performance enhancements. Our calibration method can therefore widen the applicability under various external conditions, serving as a key component for practical panorama-based machine vision systems.
Multimodal signals, including text, audio, image and video, can be integrated into Semantic Communication (SC) for providing an immersive experience with low latency and high quality at the semantic level. However, the multimodal SC has several challenges, including data heterogeneity, semantic ambiguity, and signal fading. Recent advancements in large AI models, particularly in Multimodal Language Model (MLM) and Large Language Model (LLM), offer potential solutions for these issues. To this end, we propose a Large AI Model-based Multimodal SC (LAM-MSC) framework, in which we first present the MLM-based Multimodal Alignment (MMA) that utilizes the MLM to enable the transformation between multimodal and unimodal data while preserving semantic consistency. Then, a personalized LLM-based Knowledge Base (LKB) is proposed, which allows users to perform personalized semantic extraction or recovery through the LLM. This effectively addresses the semantic ambiguity. Finally, we apply the Conditional Generative adversarial networks-based channel Estimation (CGE) to obtain Channel State Information (CSI). This approach effectively mitigates the impact of fading channels in SC. Finally, we conduct simulations that demonstrate the superior performance of the LAM-MSC framework.
Despite its clinical utility, medical image segmentation (MIS) remains a daunting task due to images' inherent complexity and variability. Vision transformers (ViTs) have recently emerged as a promising solution to improve MIS; however, they require larger training datasets than convolutional neural networks. To overcome this obstacle, data-efficient ViTs were proposed, but they are typically trained using a single source of data, which overlooks the valuable knowledge that could be leveraged from other available datasets. Naivly combining datasets from different domains can result in negative knowledge transfer (NKT), i.e., a decrease in model performance on some domains with non-negligible inter-domain heterogeneity. In this paper, we propose MDViT, the first multi-domain ViT that includes domain adapters to mitigate data-hunger and combat NKT by adaptively exploiting knowledge in multiple small data resources (domains). Further, to enhance representation learning across domains, we integrate a mutual knowledge distillation paradigm that transfers knowledge between a universal network (spanning all the domains) and auxiliary domain-specific branches. Experiments on 4 skin lesion segmentation datasets show that MDViT outperforms state-of-the-art algorithms, with superior segmentation performance and a fixed model size, at inference time, even as more domains are added. Our code is available at https://github.com/siyi-wind/MDViT.
Current state-of-the-art (SOTA) methods in visual object tracking often require extensive computational resources and vast amounts of training data, leading to a risk of overfitting. This study introduces a more efficient training strategy to mitigate overfitting and reduce computational requirements. We balance the training process with a mix of negative and positive samples from the outset, named as Joint learning with Negative samples (JN). Negative samples refer to scenarios where the object from the template is not present in the search region, which helps to prevent the model from simply memorizing the target, and instead encourages it to use the template for object location. To handle the negative samples effectively, we adopt a distribution-based head, which modeling the bounding box as distribution of distances to express uncertainty about the target's location in the presence of negative samples, offering an efficient way to manage the mixed sample training. Furthermore, our approach introduces a target-indicating token. It encapsulates the target's precise location within the template image. This method provides exact boundary details with negligible computational cost but improving performance. Our model, JN-256, exhibits superior performance on challenging benchmarks, achieving 75.8% AO on GOT-10k and 84.1% AUC on TrackingNet. Notably, JN-256 outperforms previous SOTA trackers that utilize larger models and higher input resolutions, even though it is trained with only half the number of data sampled used in those works.
The local road network information is essential for autonomous navigation. This information is commonly obtained from offline HD-Maps in terms of lane graphs. However, the local road network at a given moment can be drastically different than the one given in the offline maps; due to construction works, accidents etc. Moreover, the autonomous vehicle might be at a location not covered in the offline HD-Map. Thus, online estimation of the lane graph is crucial for widespread and reliable autonomous navigation. In this work, we tackle online Bird's-Eye-View lane graph extraction from a single onboard camera image. We propose to use prior information to increase quality of the estimations. The prior is extracted from the dataset through a transformer based Wasserstein Autoencoder. The autoencoder is then used to enhance the initial lane graph estimates. This is done through optimization of the latent space vector. The optimization encourages the lane graph estimation to be logical by discouraging it to diverge from the prior distribution. We test the method on two benchmark datasets, NuScenes and Argoverse. The results show that the proposed method significantly improves the performance compared to state-of-the-art methods.
Unsupervised Out-of-Distribution (OOD) detection consists in identifying anomalous regions in images leveraging only models trained on images of healthy anatomy. An established approach is to tokenize images and model the distribution of tokens with Auto-Regressive (AR) models. AR models are used to 1) identify anomalous tokens and 2) in-paint anomalous representations with in-distribution tokens. However, AR models are slow at inference time and prone to error accumulation issues which negatively affect OOD detection performance. Our novel method, MIM-OOD, overcomes both speed and error accumulation issues by replacing the AR model with two task-specific networks: 1) a transformer optimized to identify anomalous tokens and 2) a transformer optimized to in-paint anomalous tokens using masked image modelling (MIM). Our experiments with brain MRI anomalies show that MIM-OOD substantially outperforms AR models (DICE 0.458 vs 0.301) while achieving a nearly 25x speedup (9.5s vs 244s).
Medical image segmentation typically necessitates a large and precisely annotated dataset. However, obtaining pixel-wise annotation is a labor-intensive task that requires significant effort from domain experts, making it challenging to obtain in practical clinical scenarios. In such situations, reducing the amount of annotation required is a more practical approach. One feasible direction is sparse annotation, which involves annotating only a few slices, and has several advantages over traditional weak annotation methods such as bounding boxes and scribbles, as it preserves exact boundaries. However, learning from sparse annotation is challenging due to the scarcity of supervision signals. To address this issue, we propose a framework that can robustly learn from sparse annotation using the cross-teaching of both 3D and 2D networks. Considering the characteristic of these networks, we develop two pseudo label selection strategies, which are hard-soft confidence threshold and consistent label fusion. Our experimental results on the MMWHS dataset demonstrate that our method outperforms the state-of-the-art (SOTA) semi-supervised segmentation methods. Moreover, our approach achieves results that are comparable to the fully-supervised upper bound result.
We introduce Point-Bind, a 3D multi-modality model aligning point clouds with 2D image, language, audio, and video. Guided by ImageBind, we construct a joint embedding space between 3D and multi-modalities, enabling many promising applications, e.g., any-to-3D generation, 3D embedding arithmetic, and 3D open-world understanding. On top of this, we further present Point-LLM, the first 3D large language model (LLM) following 3D multi-modal instructions. By parameter-efficient fine-tuning techniques, Point-LLM injects the semantics of Point-Bind into pre-trained LLMs, e.g., LLaMA, which requires no 3D instruction data, but exhibits superior 3D and multi-modal question-answering capacity. We hope our work may cast a light on the community for extending 3D point clouds to multi-modality applications. Code is available at https://github.com/ZiyuGuo99/Point-Bind_Point-LLM.