Multi-source domain adaptation (MSDA) learns to predict the labels in target domain data, under the setting where all data from multiple source domains are labelled and the data from the target domain are unlabeled. To handle this problem, most of methods focus on learning invariant representations across domains. However, their success severely relies on the assumption that label distribution remains unchanged across domains. To mitigate it, we propose a new assumption, latent covariate shift, where the marginal distribution of the latent content variable changes across domains, and the conditional distribution of the label given the latent content remains invariant across domains. We introduce a latent style variable to complement the latent content variable forming a latent causal graph as the data and label generating process. We show that although the latent style variable is unidentifiable due to transitivity property in the latent space, the latent content variable can be identified up to simple scaling under some mild conditions. This motivates us to propose a novel method for MSDA, which learns the invariant label distribution conditional on the latent content variable, instead of learning invariant representations. Empirical evaluation on simulation and real data demonstrates the effectiveness of the proposed method, compared with many state-of-the-art methods based on invariant representation.
Causal representation learning exposes latent high-level causal variables behind low-level observations, which has enormous potential for a set of downstream tasks of interest. Despite this, identifying the true latent causal representation from observed data is a great challenge. In this work we focus on identifying latent causal variables. To this end, we analysis three intrinsic properties in latent space, including transitivity, permutation and scaling. We show that the transitivity severely hinders the identifiability of latent causal variables, while permutation and scaling guide the direction of identifying latent causal variable. To break the transitivity, we assume the underlying latent causal relations to be linear Gaussian models, in which the weights, mean and variance of Gaussian noise are modulated by an additionally observed variable. Under these assumptions we theoretically show that the latent causal variables can be identifiable up to trivial permutation and scaling. Built on this theoretical result, we propose a novel method, termed Structural caUsAl Variational autoEncoder, which directly learns latent causal variables, together with the mapping from the latent causal variables to the observed ones. Experimental results on synthetic and real data demonstrate the identifiable result and the ability of the proposed method for learning latent causal variables.
Noncontact particle manipulation (NPM) technology has significantly extended mankind's analysis capability into micro and nano scale, which in turn greatly promoted the development of material science and life science. Though NPM by means of electric, magnetic, and optical field has achieved great success, from the robotic perspective, it is still labor-intensive manipulation since professional human assistance is somehow mandatory in early preparation stage. Therefore, developing automated noncontact trapping of moving particles is worthwhile, particularly for applications where particle samples are rare, fragile or contact sensitive. Taking advantage of latest dynamic acoustic field modulating technology, and particularly by virtue of the great scalability of acoustic manipulation from micro-scale to sub-centimeter-scale, we propose an automated noncontact trapping of moving micro-particles with ultrasonic phased array system and microscopic vision in this paper. The main contribution of this work is for the first time, as far as we know, we achieved fully automated moving micro-particle trapping in acoustic NPM field by resorting to robotic approach. In short, the particle moving status is observed and predicted by binocular microscopic vision system, by referring to which the acoustic trapping zone is calculated and generated to capture and stably hold the particle. The problem of hand-eye relationship of noncontact robotic end-effector is also solved in this work. Experiments demonstrated the effectiveness of this work.
This paper reviews the challenge on constrained high dynamic range (HDR) imaging that was part of the New Trends in Image Restoration and Enhancement (NTIRE) workshop, held in conjunction with CVPR 2022. This manuscript focuses on the competition set-up, datasets, the proposed methods and their results. The challenge aims at estimating an HDR image from multiple respective low dynamic range (LDR) observations, which might suffer from under- or over-exposed regions and different sources of noise. The challenge is composed of two tracks with an emphasis on fidelity and complexity constraints: In Track 1, participants are asked to optimize objective fidelity scores while imposing a low-complexity constraint (i.e. solutions can not exceed a given number of operations). In Track 2, participants are asked to minimize the complexity of their solutions while imposing a constraint on fidelity scores (i.e. solutions are required to obtain a higher fidelity score than the prescribed baseline). Both tracks use the same data and metrics: Fidelity is measured by means of PSNR with respect to a ground-truth HDR image (computed both directly and with a canonical tonemapping operation), while complexity metrics include the number of Multiply-Accumulate (MAC) operations and runtime (in seconds).
In recent years, the pre-training-then-fine-tuning paradigm has yielded immense success on a wide spectrum of cross-modal tasks, such as visual question answering (VQA), in which a visual-language (VL) model is first optimized via self-supervised task objectives, e.g., masked language modeling (MLM) and image-text matching (ITM), and then fine-tuned to adapt to downstream task (e.g., VQA) via a brand-new objective function, e.g., answer prediction. The inconsistency of the objective forms not only severely limits the generalization of pre-trained VL models to downstream tasks, but also requires a large amount of labeled data for fine-tuning. To alleviate the problem, we propose an innovative VL fine-tuning paradigm (named Declaration-based Prompt Tuning, abbreviated as DPT), which jointly optimizes the objectives of pre-training and fine-tuning of VQA model, boosting the effective adaptation of pre-trained VL models to the downstream task. Specifically, DPT reformulates the objective form of VQA task via (1) textual adaptation, which converts the given questions into declarative sentence-form for prompt-tuning, and (2) task adaptation, which optimizes the objective function of VQA problem in the manner of pre-training phase. Experimental results on GQA dataset show that DPT outperforms the fine-tuned counterpart by a large margin regarding accuracy in both fully-supervised (2.68%) and zero-shot/few-shot (over 31%) settings. All the data and codes will be available to facilitate future research.
Continual Learning (CL) methods aim to enable machine learning models to learn new tasks without catastrophic forgetting of those that have been previously mastered. Existing CL approaches often keep a buffer of previously-seen samples, perform knowledge distillation, or use regularization techniques towards this goal. Despite their performance, they still suffer from interference across tasks which leads to catastrophic forgetting. To ameliorate this problem, we propose to only activate and select sparse neurons for learning current and past tasks at any stage. More parameters space and model capacity can thus be reserved for the future tasks. This minimizes the interference between parameters for different tasks. To do so, we propose a Sparse neural Network for Continual Learning (SNCL), which employs variational Bayesian sparsity priors on the activations of the neurons in all layers. Full Experience Replay (FER) provides effective supervision in learning the sparse activations of the neurons in different layers. A loss-aware reservoir-sampling strategy is developed to maintain the memory buffer. The proposed method is agnostic as to the network structures and the task boundaries. Experiments on different datasets show that our approach achieves state-of-the-art performance for mitigating forgetting.
Mammogram mass detection is crucial for diagnosing and preventing the breast cancers in clinical practice. The complementary effect of multi-view mammogram images provides valuable information about the breast anatomical prior structure and is of great significance in digital mammography interpretation. However, unlike radiologists who can utilize the natural reasoning ability to identify masses based on multiple mammographic views, how to endow the existing object detection models with the capability of multi-view reasoning is vital for decision-making in clinical diagnosis but remains the boundary to explore. In this paper, we propose an Anatomy-aware Graph convolutional Network (AGN), which is tailored for mammogram mass detection and endows existing detection methods with multi-view reasoning ability. The proposed AGN consists of three steps. Firstly, we introduce a Bipartite Graph convolutional Network (BGN) to model the intrinsic geometric and semantic relations of ipsilateral views. Secondly, considering that the visual asymmetry of bilateral views is widely adopted in clinical practice to assist the diagnosis of breast lesions, we propose an Inception Graph convolutional Network (IGN) to model the structural similarities of bilateral views. Finally, based on the constructed graphs, the multi-view information is propagated through nodes methodically, which equips the features learned from the examined view with multi-view reasoning ability. Experiments on two standard benchmarks reveal that AGN significantly exceeds the state-of-the-art performance. Visualization results show that AGN provides interpretable visual cues for clinical diagnosis.
Variational dropout (VD) is a generalization of Gaussian dropout, which aims at inferring the posterior of network weights based on a log-uniform prior on them to learn these weights as well as dropout rate simultaneously. The log-uniform prior not only interprets the regularization capacity of Gaussian dropout in network training, but also underpins the inference of such posterior. However, the log-uniform prior is an improper prior (i.e., its integral is infinite) which causes the inference of posterior to be ill-posed, thus restricting the regularization performance of VD. To address this problem, we present a new generalization of Gaussian dropout, termed variational Bayesian dropout (VBD), which turns to exploit a hierarchical prior on the network weights and infer a new joint posterior. Specifically, we implement the hierarchical prior as a zero-mean Gaussian distribution with variance sampled from a uniform hyper-prior. Then, we incorporate such a prior into inferring the joint posterior over network weights and the variance in the hierarchical prior, with which both the network training and the dropout rate estimation can be cast into a joint optimization problem. More importantly, the hierarchical prior is a proper prior which enables the inference of posterior to be well-posed. In addition, we further show that the proposed VBD can be seamlessly applied to network compression. Experiments on both classification and network compression tasks demonstrate the superior performance of the proposed VBD in terms of regularizing network training.
Minutiae extraction is of critical importance in automated fingerprint recognition. Previous works on rolled/slap fingerprints failed on latent fingerprints due to noisy ridge patterns and complex background noises. In this paper, we propose a new way to design deep convolutional network combining domain knowledge and the representation ability of deep learning. In terms of orientation estimation, segmentation, enhancement and minutiae extraction, several typical traditional methods performed well on rolled/slap fingerprints are transformed into convolutional manners and integrated as an unified plain network. We demonstrate that this pipeline is equivalent to a shallow network with fixed weights. The network is then expanded to enhance its representation ability and the weights are released to learn complex background variance from data, while preserving end-to-end differentiability. Experimental results on NIST SD27 latent database and FVC 2004 slap database demonstrate that the proposed algorithm outperforms the state-of-the-art minutiae extraction algorithms. Code is made publicly available at: https://github.com/felixTY/FingerNet.