Humans interpret scenes by recognizing both the identities and positions of objects in their observations. For a robot to perform tasks such as \enquote{pick and place}, understanding both what the objects are and where they are located is crucial. While the former has been extensively discussed in the literature that uses the large language model to enrich the text descriptions, the latter remains underexplored. In this work, we introduce the \textit{Object-Centric Instruction Augmentation (OCI)} framework to augment highly semantic and information-dense language instruction with position cues. We utilize a Multi-modal Large Language Model (MLLM) to weave knowledge of object locations into natural language instruction, thus aiding the policy network in mastering actions for versatile manipulation. Additionally, we present a feature reuse mechanism to integrate the vision-language features from off-the-shelf pre-trained MLLM into policy networks. Through a series of simulated and real-world robotic tasks, we demonstrate that robotic manipulator imitation policies trained with our enhanced instructions outperform those relying solely on traditional language instructions.
Image segmentation, real-value prediction, and cross-modal translation are critical challenges in medical imaging. In this study, we propose a versatile multi-task neural network framework, based on an enhanced Transformer U-Net architecture, capable of simultaneously, selectively, and adaptively addressing these medical image tasks. Validation is performed on a public repository of human brain MR and CT images. We decompose the traditional problem of synthesizing CT images into distinct subtasks, which include skull segmentation, Hounsfield unit (HU) value prediction, and image sequential reconstruction. To enhance the framework's versatility in handling multi-modal data, we expand the model with multiple image channels. Comparisons between synthesized CT images derived from T1-weighted and T2-Flair images were conducted, evaluating the model's capability to integrate multi-modal information from both morphological and pixel value perspectives.
The traditional definition of co-salient object detection (CoSOD) task is to segment the common salient objects in a group of relevant images. This definition is based on an assumption of group consensus consistency that is not always reasonable in the open-world setting, which results in robustness issue in the model when dealing with irrelevant images in the inputting image group under the open-word scenarios. To tackle this problem, we introduce a group selective exchange-masking (GSEM) approach for enhancing the robustness of the CoSOD model. GSEM takes two groups of images as input, each containing different types of salient objects. Based on the mixed metric we designed, GSEM selects a subset of images from each group using a novel learning-based strategy, then the selected images are exchanged. To simultaneously consider the uncertainty introduced by irrelevant images and the consensus features of the remaining relevant images in the group, we designed a latent variable generator branch and CoSOD transformer branch. The former is composed of a vector quantised-variational autoencoder to generate stochastic global variables that model uncertainty. The latter is designed to capture correlation-based local features that include group consensus. Finally, the outputs of the two branches are merged and passed to a transformer-based decoder to generate robust predictions. Taking into account that there are currently no benchmark datasets specifically designed for open-world scenarios, we constructed three open-world benchmark datasets, namely OWCoSal, OWCoSOD, and OWCoCA, based on existing datasets. By breaking the group-consistency assumption, these datasets provide effective simulations of real-world scenarios and can better evaluate the robustness and practicality of models.
Learned image compression possesses a unique challenge when incorporating non-differentiable quantization into the gradient-based training of the networks. Several quantization surrogates have been proposed to fulfill the training, but they were not systematically justified from a theoretical perspective. We fill this gap by contrasting uniform scalar quantization, the most widely used category with rounding being its simplest case, and its training surrogates. In principle, we find two factors crucial: one is the discrepancy between the surrogate and rounding, leading to train-test mismatch; the other is gradient estimation risk due to the surrogate, which consists of bias and variance of the gradient estimation. Our analyses and simulations imply that there is a tradeoff between the train-test mismatch and the gradient estimation risk, and the tradeoff varies across different network structures. Motivated by these analyses, we present a method based on stochastic uniform annealing, which has an adjustable temperature coefficient to control the tradeoff. Moreover, our analyses enlighten us as to two subtle tricks: one is to set an appropriate lower bound for the variance parameter of the estimated quantized latent distribution, which effectively reduces the train-test mismatch; the other is to use zero-center quantization with partial stop-gradient, which reduces the gradient estimation variance and thus stabilize the training. Our method with the tricks is verified to outperform the existing practices of quantization surrogates on a variety of representative image compression networks.
Video analysis tasks rely heavily on identifying the pixels from different frames that correspond to the same visual target. To tackle this problem, recent studies have advocated feature learning methods that aim to learn distinctive representations to match the pixels, especially in a self-supervised fashion. Unfortunately, these methods have difficulties for tiny or even single-pixel visual targets. Pixel-wise video correspondences were traditionally related to optical flows, which however lead to deterministic correspondences and lack robustness on real-world videos. We address the problem of learning features for establishing pixel-wise correspondences. Motivated by optical flows as well as the self-supervised feature learning, we propose to use not only labeled synthetic videos but also unlabeled real-world videos for learning fine-grained representations in a holistic framework. We adopt an adversarial learning scheme to enhance the generalization ability of the learned features. Moreover, we design a coarse-to-fine framework to pursue high computational efficiency. Our experimental results on a series of correspondence-based tasks demonstrate that the proposed method outperforms state-of-the-art rivals in both accuracy and efficiency.
Estimating 6D poses and reconstructing 3D shapes of objects in open-world scenes from RGB-depth image pairs is challenging. Many existing methods rely on learning geometric features that correspond to specific templates while disregarding shape variations and pose differences among objects in the same category. As a result, these methods underperform when handling unseen object instances in complex environments. In contrast, other approaches aim to achieve category-level estimation and reconstruction by leveraging normalized geometric structure priors, but the static prior-based reconstruction struggles with substantial intra-class variations. To solve these problems, we propose the DTF-Net, a novel framework for pose estimation and shape reconstruction based on implicit neural fields of object categories. In DTF-Net, we design a deformable template field to represent the general category-wise shape latent features and intra-category geometric deformation features. The field establishes continuous shape correspondences, deforming the category template into arbitrary observed instances to accomplish shape reconstruction. We introduce a pose regression module that shares the deformation features and template codes from the fields to estimate the accurate 6D pose of each object in the scene. We integrate a multi-modal representation extraction module to extract object features and semantic masks, enabling end-to-end inference. Moreover, during training, we implement a shape-invariant training strategy and a viewpoint sampling method to further enhance the model's capability to extract object pose features. Extensive experiments on the REAL275 and CAMERA25 datasets demonstrate the superiority of DTF-Net in both synthetic and real scenes. Furthermore, we show that DTF-Net effectively supports grasping tasks with a real robot arm.
Several recent studies advocate the use of spectral discriminators, which evaluate the Fourier spectra of images for generative modeling. However, the effectiveness of the spectral discriminators is not well interpreted yet. We tackle this issue by examining the spectral discriminators in the context of perceptual image super-resolution (i.e., GAN-based SR), as SR image quality is susceptible to spectral changes. Our analyses reveal that the spectral discriminator indeed performs better than the ordinary (a.k.a. spatial) discriminator in identifying the differences in the high-frequency range; however, the spatial discriminator holds an advantage in the low-frequency range. Thus, we suggest that the spectral and spatial discriminators shall be used simultaneously. Moreover, we improve the spectral discriminators by first calculating the patch-wise Fourier spectrum and then aggregating the spectra by Transformer. We verify the effectiveness of the proposed method twofold. On the one hand, thanks to the additional spectral discriminator, our obtained SR images have their spectra better aligned to those of the real images, which leads to a better PD tradeoff. On the other hand, our ensembled discriminator predicts the perceptual quality more accurately, as evidenced in the no-reference image quality assessment task.
Video compression relies heavily on exploiting the temporal redundancy between video frames, which is usually achieved by estimating and using the motion information. The motion information is represented as optical flows in most of the existing deep video compression networks. Indeed, these networks often adopt pre-trained optical flow estimation networks for motion estimation. The optical flows, however, may be less suitable for video compression due to the following two factors. First, the optical flow estimation networks were trained to perform inter-frame prediction as accurately as possible, but the optical flows themselves may cost too many bits to encode. Second, the optical flow estimation networks were trained on synthetic data, and may not generalize well enough to real-world videos. We address the twofold limitations by enhancing the optical flows in two stages: offline and online. In the offline stage, we fine-tune a trained optical flow estimation network with the motion information provided by a traditional (non-deep) video compression scheme, e.g. H.266/VVC, as we believe the motion information of H.266/VVC achieves a better rate-distortion trade-off. In the online stage, we further optimize the latent features of the optical flows with a gradient descent-based algorithm for the video to be compressed, so as to enhance the adaptivity of the optical flows. We conduct experiments on a state-of-the-art deep video compression scheme, DCVC. Experimental results demonstrate that the proposed offline and online enhancement together achieves on average 12.8% bitrate saving on the tested videos, without increasing the model or computational complexity of the decoder side.
Almost all digital videos are coded into compact representations before being transmitted. Such compact representations need to be decoded back to pixels before being displayed to human and - as usual - before being processed/analyzed by machine vision algorithms. For machine vision, it is more efficient at least conceptually, to process/analyze the coded representations directly without decoding them into pixels. Motivated by this concept, we propose a learned versatile video coding (LVVC) framework, which targets on learning compact representations to support both decoding and direct processing/analysis, thereby being versatile for both human and machine vision. Our LVVC framework has a feature-based compression loop, where one frame is encoded (resp. decoded) to intermediate features, and the intermediate features are referenced for encoding (resp. decoding) the following frames. Our proposed feature-based compression loop has two key technologies, one is feature-based temporal context mining, and the other is cross-domain motion encoder/decoder. With the LVVC framework, the intermediate features may be used to reconstruct videos, or be fed into different task networks. The LVVC framework is implemented and evaluated with video reconstruction, video processing, and video analysis tasks on the well-established benchmark datasets. The evaluation results demonstrate the compression efficiency of the proposed LVVC framework.
One tough problem of image inpainting is to restore complex structures in the corrupted regions. It motivates interactive image inpainting which leverages additional hints, e.g., sketches, to assist the inpainting process. Sketch is simple and intuitive to end users, but meanwhile has free forms with much randomness. Such randomness may confuse the inpainting models, and incur severe artifacts in completed images. To address this problem, we propose a two-stage image inpainting method termed SketchRefiner. In the first stage, we propose using a cross-correlation loss function to robustly calibrate and refine the user-provided sketches in a coarse-to-fine fashion. In the second stage, we learn to extract informative features from the abstracted sketches in the feature space and modulate the inpainting process. We also propose an algorithm to simulate real sketches automatically and build a test protocol with different applications. Experimental results on public datasets demonstrate that SketchRefiner effectively utilizes sketch information and eliminates the artifacts due to the free-form sketches. Our method consistently outperforms the state-of-the-art ones both qualitatively and quantitatively, meanwhile revealing great potential in real-world applications. Our code and dataset are available.