Theoretical properties of bilevel problems are well studied when the lower-level problem is strongly convex. In this work, we focus on bilevel optimization problems without the strong-convexity assumption. In these cases, we first show that the common local optimality measures such as KKT condition or regularization can lead to undesired consequences. Then, we aim to identify the mildest conditions that make bilevel problems tractable. We identify two classes of growth conditions on the lower-level objective that leads to continuity. Under these assumptions, we show that the local optimality of the bilevel problem can be defined via the Goldstein stationarity condition of the hyper-objective. We then propose the Inexact Gradient-Free Method (IGFM) to solve the bilevel problem, using an approximate zeroth order oracle that is of independent interest. Our non-asymptotic analysis demonstrates that the proposed method can find a $(\delta, \varepsilon)$ Goldstein stationary point for bilevel problems with a zeroth order oracle complexity that is polynomial in $d, 1/\delta$ and $1/\varepsilon$.

Stance detection refers to the task of extracting the standpoint (Favor, Against or Neither) towards a target in given texts. Such research gains increasing attention with the proliferation of social media contents. The conventional framework of handling stance detection is converting it into text classification tasks. Deep learning models have already replaced rule-based models and traditional machine learning models in solving such problems. Current deep neural networks are facing two main challenges which are insufficient labeled data and information in social media posts and the unexplainable nature of deep learning models. A new pre-trained language model chatGPT was launched on Nov 30, 2022. For the stance detection tasks, our experiments show that ChatGPT can achieve SOTA or similar performance for commonly used datasets including SemEval-2016 and P-Stance. At the same time, ChatGPT can provide explanation for its own prediction, which is beyond the capability of any existing model. The explanations for the cases it cannot provide classification results are especially useful. ChatGPT has the potential to be the best AI model for stance detection tasks in NLP, or at least change the research paradigm of this field. ChatGPT also opens up the possibility of building explanatory AI for stance detection.

Attention-based neural networks, such as Transformers, have become ubiquitous in numerous applications, including computer vision, natural language processing, and time-series analysis. In all kinds of attention networks, the attention maps are crucial as they encode semantic dependencies between input tokens. However, most existing attention networks perform modeling or reasoning based on representations, wherein the attention maps of different layers are learned separately without explicit interactions. In this paper, we propose a novel and generic evolving attention mechanism, which directly models the evolution of inter-token relationships through a chain of residual convolutional modules. The major motivations are twofold. On the one hand, the attention maps in different layers share transferable knowledge, thus adding a residual connection can facilitate the information flow of inter-token relationships across layers. On the other hand, there is naturally an evolutionary trend among attention maps at different abstraction levels, so it is beneficial to exploit a dedicated convolution-based module to capture this process. Equipped with the proposed mechanism, the convolution-enhanced evolving attention networks achieve superior performance in various applications, including time-series representation, natural language understanding, machine translation, and image classification. Especially on time-series representation tasks, Evolving Attention-enhanced Dilated Convolutional (EA-DC-) Transformer outperforms state-of-the-art models significantly, achieving an average of 17% improvement compared to the best SOTA. To the best of our knowledge, this is the first work that explicitly models the layer-wise evolution of attention maps. Our implementation is available at https://github.com/pkuyym/EvolvingAttention

Ultrasound is progressing toward becoming an affordable and versatile solution to medical imaging. With the advent of COVID-19 global pandemic, there is a need to fully automate ultrasound imaging as it requires trained operators in close proximity to patients for long period of time. In this work, we investigate the important yet seldom-studied problem of scan target localization, under the setting of lung ultrasound imaging. We propose a purely vision-based, data driven method that incorporates learning-based computer vision techniques. We combine a human pose estimation model with a specially designed regression model to predict the lung ultrasound scan targets, and deploy multiview stereo vision to enhance the consistency of 3D target localization. While related works mostly focus on phantom experiments, we collect data from 30 human subjects for testing. Our method attains an accuracy level of 15.52 (9.47) mm for probe positioning and 4.32 (3.69){\deg} for probe orientation, with a success rate above 80% under an error threshold of 25mm for all scan targets. Moreover, our approach can serve as a general solution to other types of ultrasound modalities. The code for implementation has been released.

Directly training a document-to-document (Doc2Doc) neural machine translation (NMT) via Transformer from scratch, especially on small datasets usually fails to converge. Our dedicated probing tasks show that 1) both the absolute position and relative position information gets gradually weakened or even vanished once it reaches the upper encoder layers, and 2) the vanishing of absolute position information in encoder output causes the training failure of Doc2Doc NMT. To alleviate this problem, we propose a position-aware Transformer (P-Transformer) to enhance both the absolute and relative position information in both self-attention and cross-attention. Specifically, we integrate absolute positional information, i.e., position embeddings, into the query-key pairs both in self-attention and cross-attention through a simple yet effective addition operation. Moreover, we also integrate relative position encoding in self-attention. The proposed P-Transformer utilizes sinusoidal position encoding and does not require any task-specified position embedding, segment embedding, or attention mechanism. Through the above methods, we build a Doc2Doc NMT model with P-Transformer, which ingests the source document and completely generates the target document in a sequence-to-sequence (seq2seq) way. In addition, P-Transformer can be applied to seq2seq-based document-to-sentence (Doc2Sent) and sentence-to-sentence (Sent2Sent) translation. Extensive experimental results of Doc2Doc NMT show that P-Transformer significantly outperforms strong baselines on widely-used 9 document-level datasets in 7 language pairs, covering small-, middle-, and large-scales, and achieves a new state-of-the-art. Experimentation on discourse phenomena shows that our Doc2Doc NMT models improve the translation quality in both BLEU and discourse coherence. We make our code available on Github.

Semantic Change Detection (SCD) refers to the task of simultaneously extracting the changed areas and the semantic categories (before and after the changes) in Remote Sensing Images (RSIs). This is more meaningful than Binary Change Detection (BCD) since it enables detailed change analysis in the observed areas. Previous works established triple-branch Convolutional Neural Network (CNN) architectures as the paradigm for SCD. However, it remains challenging to exploit semantic information with a limited amount of change samples. In this work, we investigate to jointly consider the spatio-temporal dependencies to improve the accuracy of SCD. First, we propose a SCanFormer (Semantic Change Transformer) to explicitly model the 'from-to' semantic transitions between the bi-temporal RSIs. Then, we introduce a semantic learning scheme to leverage the spatio-temporal constraints, which are coherent to the SCD task, to guide the learning of semantic changes. The resulting network (ScanNet) significantly outperforms the baseline method in terms of both detection of critical semantic changes and semantic consistency in the obtained bi-temporal results. It achieves the SOTA accuracy on two benchmark datasets for the SCD.

In this paper, we show the surprisingly good properties of plain vision transformers for body pose estimation from various aspects, namely simplicity in model structure, scalability in model size, flexibility in training paradigm, and transferability of knowledge between models, through a simple baseline model dubbed ViTPose. Specifically, ViTPose employs the plain and non-hierarchical vision transformer as an encoder to encode features and a lightweight decoder to decode body keypoints in either a top-down or a bottom-up manner. It can be scaled up from about 20M to 1B parameters by taking advantage of the scalable model capacity and high parallelism of the vision transformer, setting a new Pareto front for throughput and performance. Besides, ViTPose is very flexible regarding the attention type, input resolution, and pre-training and fine-tuning strategy. Based on the flexibility, a novel ViTPose+ model is proposed to deal with heterogeneous body keypoint categories in different types of body pose estimation tasks via knowledge factorization, i.e., adopting task-agnostic and task-specific feed-forward networks in the transformer. We also empirically demonstrate that the knowledge of large ViTPose models can be easily transferred to small ones via a simple knowledge token. Experimental results show that our ViTPose model outperforms representative methods on the challenging MS COCO Human Keypoint Detection benchmark at both top-down and bottom-up settings. Furthermore, our ViTPose+ model achieves state-of-the-art performance simultaneously on a series of body pose estimation tasks, including MS COCO, AI Challenger, OCHuman, MPII for human keypoint detection, COCO-Wholebody for whole-body keypoint detection, as well as AP-10K and APT-36K for animal keypoint detection, without sacrificing inference speed.

Considerable progress has recently been made in leveraging CLIP (Contrastive Language-Image Pre-Training) models for text-guided image manipulation. However, all existing works rely on additional generative models to ensure the quality of results, because CLIP alone cannot provide enough guidance information for fine-scale pixel-level changes. In this paper, we introduce CLIPVG, a text-guided image manipulation framework using differentiable vector graphics, which is also the first CLIP-based general image manipulation framework that does not require any additional generative models. We demonstrate that CLIPVG can not only achieve state-of-art performance in both semantic correctness and synthesis quality, but also is flexible enough to support various applications far beyond the capability of all existing methods.

Recently, the joint learning framework (JOINT) integrates matching based transductive reasoning and online inductive learning to achieve accurate and robust semi-supervised video object segmentation (SVOS). However, using the mask embedding as the label to guide the generation of target features in the two branches may result in inadequate target representation and degrade the performance. Besides, how to reasonably fuse the target features in the two different branches rather than simply adding them together to avoid the adverse effect of one dominant branch has not been investigated. In this paper, we propose a novel framework that emphasizes Learning to Learn Better (LLB) target features for SVOS, termed LLB, where we design the discriminative label generation module (DLGM) and the adaptive fusion module to address these issues. Technically, the DLGM takes the background-filtered frame instead of the target mask as input and adopts a lightweight encoder to generate the target features, which serves as the label of the online few-shot learner and the value of the decoder in the transformer to guide the two branches to learn more discriminative target representation. The adaptive fusion module maintains a learnable gate for each branch, which reweighs the element-wise feature representation and allows an adaptive amount of target information in each branch flowing to the fused target feature, thus preventing one branch from being dominant and making the target feature more robust to distractor. Extensive experiments on public benchmarks show that our proposed LLB method achieves state-of-the-art performance.

Clustering has been extensively studied in centralized settings, but relatively unexplored in federated ones that data are distributed among multiple clients and can only be kept local at the clients. The necessity to invest more resources in improving federated clustering methods is twofold: 1) The performance of supervised federated learning models can benefit from clustering. 2) It is non-trivial to extend centralized ones to perform federated clustering tasks. In centralized settings, various deep clustering methods that perform dimensionality reduction and clustering jointly have achieved great success. To obtain high-quality cluster information, it is natural but non-trivial to extend these methods to federated settings. For this purpose, we propose a simple but effective federated deep clustering method. It requires only one communication round between the central server and clients, can run asynchronously, and can handle device failures. Moreover, although most studies have highlighted adverse effects of the non-independent and identically distributed (non-IID) data across clients, experimental results indicate that the proposed method can significantly benefit from this scenario.