4Real: Towards Photorealistic 4D Scene Generation via Video Diffusion Models

Existing dynamic scene generation methods mostly rely on distilling knowledge from pre-trained 3D generative models, which are typically fine-tuned on synthetic object datasets. As a result, the generated scenes are often object-centric and lack photorealism. To address these limitations, we introduce a novel pipeline designed for photorealistic text-to-4D scene generation, discarding the dependency on multi-view generative models and instead fully utilizing video generative models trained on diverse real-world datasets. Our method begins by generating a reference video using the video generation model. We then learn the canonical 3D representation of the video using a freeze-time video, delicately generated from the reference video. To handle inconsistencies in the freeze-time video, we jointly learn a per-frame deformation to model these imperfections. We then learn the temporal deformation based on the canonical representation to capture dynamic interactions in the reference video. The pipeline facilitates the generation of dynamic scenes with enhanced photorealism and structural integrity, viewable from multiple perspectives, thereby setting a new standard in 4D scene generation.

WeightedPose: Generalizable Cross-Pose Estimation via Weighted SVD

We present a novel method for robotic manipulation tasks in human environments that require reasoning about the 3D geometric relationship between a pair of objects. Traditional end-to-end trained policies, which map from pixel observations to low-level robot actions, struggle to reason about complex pose relationships and have difficulty generalizing to unseen object configurations. To address these challenges, we propose a method that learns to reason about the 3D geometric relationship between objects, focusing on the relationship between key parts on one object with respect to key parts on another object. Our standalone model utilizes Weighted SVD to reason about both pose relationships between articulated parts and between free-floating objects. This approach allows the robot to understand the relationship between the oven door and the oven body, as well as the relationship between the lasagna plate and the oven, for example. By considering the 3D geometric relationship between objects, our method enables robots to perform complex manipulation tasks that reason about object-centric representations. We open source the code and demonstrate the results here

CoGS: Controllable Gaussian Splatting

Capturing and re-animating the 3D structure of articulated objects present significant barriers. On one hand, methods requiring extensively calibrated multi-view setups are prohibitively complex and resource-intensive, limiting their practical applicability. On the other hand, while single-camera Neural Radiance Fields (NeRFs) offer a more streamlined approach, they have excessive training and rendering costs. 3D Gaussian Splatting would be a suitable alternative but for two reasons. Firstly, existing methods for 3D dynamic Gaussians require synchronized multi-view cameras, and secondly, the lack of controllability in dynamic scenarios. We present CoGS, a method for Controllable Gaussian Splatting, that enables the direct manipulation of scene elements, offering real-time control of dynamic scenes without the prerequisite of pre-computing control signals. We evaluated CoGS using both synthetic and real-world datasets that include dynamic objects that differ in degree of difficulty. In our evaluations, CoGS consistently outperformed existing dynamic and controllable neural representations in terms of visual fidelity.

SubZero: Subspace Zero-Shot MRI Reconstruction

Recently introduced zero-shot self-supervised learning (ZS-SSL) has shown potential in accelerated MRI in a scan-specific scenario, which enabled high-quality reconstructions without access to a large training dataset. ZS-SSL has been further combined with the subspace model to accelerate 2D T2-shuffling acquisitions. In this work, we propose a parallel network framework and introduce an attention mechanism to improve subspace-based zero-shot self-supervised learning and enable higher acceleration factors. We name our method SubZero and demonstrate that it can achieve improved performance compared with current methods in T1 and T2 mapping acquisitions.

Instruct2Attack: Language-Guided Semantic Adversarial Attacks

We propose Instruct2Attack (I2A), a language-guided semantic attack that generates semantically meaningful perturbations according to free-form language instructions. We make use of state-of-the-art latent diffusion models, where we adversarially guide the reverse diffusion process to search for an adversarial latent code conditioned on the input image and text instruction. Compared to existing noise-based and semantic attacks, I2A generates more natural and diverse adversarial examples while providing better controllability and interpretability. We further automate the attack process with GPT-4 to generate diverse image-specific text instructions. We show that I2A can successfully break state-of-the-art deep neural networks even under strong adversarial defenses, and demonstrate great transferability among a variety of network architectures.

Unsupervised Style-based Explicit 3D Face Reconstruction from Single Image

Inferring 3D object structures from a single image is an ill-posed task due to depth ambiguity and occlusion. Typical resolutions in the literature include leveraging 2D or 3D ground truth for supervised learning, as well as imposing hand-crafted symmetry priors or using an implicit representation to hallucinate novel viewpoints for unsupervised methods. In this work, we propose a general adversarial learning framework for solving Unsupervised 2D to Explicit 3D Style Transfer (UE3DST). Specifically, we merge two architectures: the unsupervised explicit 3D reconstruction network of Wu et al.\ and the Generative Adversarial Network (GAN) named StarGAN-v2. We experiment across three facial datasets (Basel Face Model, 3DFAW and CelebA-HQ) and show that our solution is able to outperform well established solutions such as DepthNet in 3D reconstruction and Pix2NeRF in conditional style transfer, while we also justify the individual contributions of our model components via ablation. In contrast to the aforementioned baselines, our scheme produces features for explicit 3D rendering, which can be manipulated and utilized in downstream tasks.

DyLiN: Making Light Field Networks Dynamic

Light Field Networks, the re-formulations of radiance fields to oriented rays, are magnitudes faster than their coordinate network counterparts, and provide higher fidelity with respect to representing 3D structures from 2D observations. They would be well suited for generic scene representation and manipulation, but suffer from one problem: they are limited to holistic and static scenes. In this paper, we propose the Dynamic Light Field Network (DyLiN) method that can handle non-rigid deformations, including topological changes. We learn a deformation field from input rays to canonical rays, and lift them into a higher dimensional space to handle discontinuities. We further introduce CoDyLiN, which augments DyLiN with controllable attribute inputs. We train both models via knowledge distillation from pretrained dynamic radiance fields. We evaluated DyLiN using both synthetic and real world datasets that include various non-rigid deformations. DyLiN qualitatively outperformed and quantitatively matched state-of-the-art methods in terms of visual fidelity, while being 25 - 71x computationally faster. We also tested CoDyLiN on attribute annotated data and it surpassed its teacher model. Project page: https://dylin2023.github.io .

Patch Network for medical image Segmentation

Accurate and fast segmentation of medical images is clinically essential, yet current research methods include convolutional neural networks with fast inference speed but difficulty in learning image contextual features, and transformer with good performance but high hardware requirements. In this paper, we present a Patch Network (PNet) that incorporates the Swin Transformer notion into a convolutional neural network, allowing it to gather richer contextual information while achieving the balance of speed and accuracy. We test our PNet on Polyp(CVC-ClinicDB and ETIS- LaribPolypDB), Skin(ISIC-2018 Skin lesion segmentation challenge dataset) segmentation datasets. Our PNet achieves SOTA performance in both speed and accuracy.

Non-pooling Network for medical image segmentation

Existing studies tend tofocus onmodel modifications and integration with higher accuracy, which improve performance but also carry huge computational costs, resulting in longer detection times. Inmedical imaging, the use of time is extremely sensitive. And at present most of the semantic segmentation models have encoder-decoder structure or double branch structure. Their several times of the pooling use with high-level semantic information extraction operation cause information loss although there si a reverse pooling or other similar action to restore information loss of pooling operation. In addition, we notice that visual attention mechanism has superior performance on a variety of tasks. Given this, this paper proposes non-pooling network(NPNet), non-pooling commendably reduces the loss of information and attention enhancement m o d u l e ( A M ) effectively increases the weight of useful information. The method greatly reduces the number of parametersand computation costs by the shallow neural network structure. We evaluate the semantic segmentation model of our NPNet on three benchmark datasets comparing w i t h multiple current state-of-the-art(SOTA) models, and the implementation results show thatour NPNetachieves SOTA performance, with an excellent balance between accuracyand speed.

CoNFies: Controllable Neural Face Avatars

Neural Radiance Fields (NeRF) are compelling techniques for modeling dynamic 3D scenes from 2D image collections. These volumetric representations would be well suited for synthesizing novel facial expressions but for two problems. First, deformable NeRFs are object agnostic and model holistic movement of the scene: they can replay how the motion changes over time, but they cannot alter it in an interpretable way. Second, controllable volumetric representations typically require either time-consuming manual annotations or 3D supervision to provide semantic meaning to the scene. We propose a controllable neural representation for face self-portraits (CoNFies), that solves both of these problems within a common framework, and it can rely on automated processing. We use automated facial action recognition (AFAR) to characterize facial expressions as a combination of action units (AU) and their intensities. AUs provide both the semantic locations and control labels for the system. CoNFies outperformed competing methods for novel view and expression synthesis in terms of visual and anatomic fidelity of expressions.