Controllable image semantic understanding tasks, such as captioning or segmentation, necessitate users to input a prompt (e.g., text or bounding boxes) to predict a unique outcome, presenting challenges such as high-cost prompt input or limited information output. This paper introduces a new task ``Image Collaborative Segmentation and Captioning'' (SegCaptioning), which aims to translate a straightforward prompt, like a bounding box around an object, into diverse semantic interpretations represented by (caption, masks) pairs, allowing flexible result selection by users. This task poses significant challenges, including accurately capturing a user's intention from a minimal prompt while simultaneously predicting multiple semantically aligned caption words and masks. Technically, we propose a novel Scene Graph Guided Diffusion Model that leverages structured scene graph features for correlated mask-caption prediction. Initially, we introduce a Prompt-Centric Scene Graph Adaptor to map a user's prompt to a scene graph, effectively capturing his intention. Subsequently, we employ a diffusion process incorporating a Scene Graph Guided Bimodal Transformer to predict correlated caption-mask pairs by uncovering intricate correlations between them. To ensure accurate alignment, we design a Multi-Entities Contrastive Learning loss to explicitly align visual and textual entities by considering inter-modal similarity, resulting in well-aligned caption-mask pairs. Extensive experiments conducted on two datasets demonstrate that SGDiff achieves superior performance in SegCaptioning, yielding promising results for both captioning and segmentation tasks with minimal prompt input.

Retrieval-augmented generation (RAG) systems empower large language models (LLMs) to access external knowledge during inference.Recent advances have enabled LLMs to act as search agents via reinforcement learning (RL), improving information acquisition through multi-turn interactions with retrieval engines.However, existing approaches either optimize retrieval using search-only metrics (e.g., NDCG) that ignore downstream utility or fine-tune the entire LLM to jointly reason and retrieve-entangling retrieval with generation and limiting the real search utility and compatibility with frozen or proprietary models.In this work, we propose s3, a lightweight, modelagnostic framework that decouples the searcher from the generator and trains the searcher using a Gain Beyond RAG reward: the improvement in generation accuracy over naïve RAG.s3 requires only 2.4k training samples to outperform baselines trained on over 70× more data, consistently delivering stronger downstream performance across six general QA and five medical QA benchmarks. 1

Interactive image segmentation (IIS) has emerged as a promising technique for decreasing annotation time. Substantial progress has been made in pre- and post-processing for IIS, but the critical issue of interaction ambiguity, notably hindering segmentation quality, has been under-researched. To address this, we introduce ADAPTIVE CLICK - a click-aware transformer incorporating an adaptive focal loss (AFL) that tackles annotation inconsistencies with tools for mask- and pixel-level ambiguity resolution. To the best of our knowledge, AdaptiveClick is the first transformer-based, mask-adaptive segmentation framework for IIS. The key ingredient of our method is the click-aware mask-adaptive transformer decoder (CAMD), which enhances the interaction between click and image features. Additionally, AdaptiveClick enables pixel-adaptive differentiation of hard and easy samples in the decision space, independent of their varying distributions. This is primarily achieved by optimizing a generalized AFL with a theoretical guarantee, where two adaptive coefficients control the ratio of gradient values for hard and easy pixels. Our analysis reveals that the commonly used Focal and BCE losses can be considered special cases of the proposed AFL. With a plain ViT backbone, extensive experimental results on nine datasets demonstrate the superiority of AdaptiveClick compared to state-of-the-art methods. The source code is publicly available at https://github.com/lab206/AdaptiveClick.

Recent advancements have introduced machine learning frameworks to enhance the Branch and Bound (B\&B) branching policies for solving Mixed Integer Linear Programming (MILP). These methods, primarily relying on imitation learning of Strong Branching, have shown superior performance. However, collecting expert samples for imitation learning, particularly for Strong Branching, is a time-consuming endeavor. To address this challenge, we propose \textbf{C}ontrastive Learning with \textbf{A}ugmented \textbf{M}ILPs for \textbf{Branch}ing (CAMBranch), a framework that generates Augmented MILPs (AMILPs) by applying variable shifting to limited expert data from their original MILPs. This approach enables the acquisition of a considerable number of labeled expert samples. CAMBranch leverages both MILPs and AMILPs for imitation learning and employs contrastive learning to enhance the model's ability to capture MILP features, thereby improving the quality of branching decisions. Experimental results demonstrate that CAMBranch, trained with only 10\% of the complete dataset, exhibits superior performance. Ablation studies further validate the effectiveness of our method.

LiDAR-based Moving Object Segmentation (MOS) aims to locate and segment moving objects in point clouds of the current scan using motion information from previous scans. Despite the promising results achieved by previous MOS methods, several key issues, such as the weak coupling of temporal and spatial information, still need further study. In this paper, we propose a novel LiDAR-based 3D Moving Object Segmentation with Motion-aware State Space Model, termed MambaMOS. Firstly, we develop a novel embedding module, the Time Clue Bootstrapping Embedding (TCBE), to enhance the coupling of temporal and spatial information in point clouds and alleviate the issue of overlooked temporal clues. Secondly, we introduce the Motion-aware State Space Model (MSSM) to endow the model with the capacity to understand the temporal correlations of the same object across different time steps. Specifically, MSSM emphasizes the motion states of the same object at different time steps through two distinct temporal modeling and correlation steps. We utilize an improved state space model to represent these motion differences, significantly modeling the motion states. Finally, extensive experiments on the SemanticKITTI-MOS and KITTI-Road benchmarks demonstrate that the proposed MambaMOS achieves state-of-the-art performance. The source code is publicly available at https://github.com/Terminal-K/MambaMOS

Pengcheng Jiang, Jiacheng Lin, Zifeng Wang, Jimeng Sun, Jiawei Han. Proceedings of the 2024 Conference of the North American Chapter of the Association for Computational Linguistics: Human Language Technologies (Volume 1: Long Papers). 2024.

Introduced to enhance the efficiency of large language model (LLM) inference, speculative decoding operates by having a smaller model generate a draft. A larger target model then reviews this draft to align with its output, and any acceptance by the target model results in a reduction of the number of the target model runs, ultimately improving efficiency. However, the drafting process in speculative decoding includes slow autoregressive generation and allocates equal time to generating tokens, irrespective of their importance. These inefficiencies collectively contribute to the suboptimal performance of speculative decoding. To further improve LLM inference, we introduce Cascade Speculative Drafting (CS Drafting), a speculative execution algorithm that incorporates two types of cascades. The *Vertical Cascade* eliminates autoregressive generation from neural models, while the *Horizontal Cascade* optimizes time allocation in drafting for improved efficiency. Combining both cascades, CS Drafting achieves greater speedup compared to the baselines in our experiments, while preserving the same output distribution as the target model. Our code is publicly available at https://github.com/lfsszd/CS-Drafting.

Abstract Although social robots equipped with visual devices may leak user information, countermeasures for ensuring privacy are not readily available, making visual privacy protection problematic. In this article, a semi‐supervised learning algorithm is proposed for visual privacy behaviour recognition based on an improved generative adversarial network for social robots; it is called PBR‐GAN. A 9‐layer residual generator network enhances the data quality, and a 10‐layer discriminator network strengthens the feature extraction. A tailored objective function, loss function, and strategy are proposed to dynamically adjust the learning rate to guarantee high performance. A social robot platform and architecture for visual privacy recognition and protection are implemented. The recognition accuracy of the proposed PBR‐GAN is compared with Inception_v3, SS‐GAN, and SF‐GAN. The average recognition accuracy of the proposed PBR‐GAN is 85.91%, which is improved by 3.93%, 9.91%, and 1.73% compared with the performance of Inception_v3, SS‐GAN, and SF‐GAN respectively. Through a case study, seven situations are considered related to privacy at home, and develop training and test datasets with 8,720 and 1,280 images, respectively, are developed. The proposed PBR‐GAN recognises the designed visual privacy information with an average accuracy of 89.91%.

The electrocardiogram (ECG) is a highly cost-effective and convenient diagnostic tool that can aid in the diagnosis of a wide range of cardiovascular conditions, such as arrhythmias , myocardial ischemia , myocardial infarction, and heart failure. Its usefulness lies in its ability to provide information about the electrical activity and rhythm of the heart, making it an essential component of the diagnostic process for many cardiac conditions. However, the interpretation of ECG signals often requires a high level of expertise from medical professionals. When reading complex ECGs, noncardiologists often need to consult with cardiologists. Even cardiologists may make errors in their judgments after prolonged reading of ECGs. Therefore, the accurate and timely diagnosis of cardiovascular diseases using ECG signals is a complex task. In this paper, we categorize this problem as a multilabel, multidimensional time-series data anomaly detection task. We propose a method called T w o 3 -AnoECG for ECG anomaly detection, which improves upon existing ECG feature extraction and data segmentation methods and introduces a two-stream network that is trained in two stages using two double-throw switches to control the input and model structure of each stage. We further propose T w o 3 -EnsECG, which combines the anomaly score generated by T w o 3 -AnoECG and multiple baseline methods, to improve the overall performance of anomaly detection in ECG signals. The experimental results demonstrate the effectiveness of our proposed method.

Compositional Zero-Shot Learning (CZSL) has been applied to various scenarios, including scene understanding, visual-language representation, and domain adaptation. Despite numerous endeavours and significant advancements, the crucial issues of fuzzy conceptualization of visual attributes and insufficient inter-class connectivity, have remained insufficiently addressed. To address these issues, we propose Learning Visual Attributes Representation for Compositional Zero-Shot Learning (LVAR-CZSL), which has the ability to learn visual attributes and inter-class dependencies. LVAR-CZSL is mainly composed of two key components: the Visual Attribute Representation Module (VARM) and the Connected Learning Module (CLM). Specifically, VARM extracts detailed attributes and object visual features from global visual features, resolving the issue of fuzzy visual attribute concepts. Moreover, CLM endows LVAR-CZSL with the capability to perceive connectivity between different attributes and objects, effectively enhancing inter-class connectivity. To establish a close connection between VARM and CLM and minimize the gap between image and text features, we introduce the composition-attribute-object Joint Scoring Function (JSF). Additionally, we propose Joint Loss Function (JLF) to optimize the learning process of VARM and CLM. The experiment results on four datasets show that LVAR-CZSL achieves state-of-the-art performance. The code is available athttps://github.com/mxjmxj1/LVAR-CZSL.

Interactive image segmentation (IIS) has been widely used in various fields, such as medicine, industry, etc. However, some core issues, such as pixel imbalance, remain unresolved so far. Different from existing methods based on pre-processing or post-processing, we analyze the cause of pixel imbalance in depth from the two perspectives of pixel number and pixel difficulty. Based on this, a novel and unified Click-pixel Cognition Fusion network with Balanced Cut (CCF-BC) is proposed in this paper. On the one hand, the Click-pixel Cognition Fusion (CCF) module, inspired by the human cognition mechanism, is designed to increase the number of click-related pixels (namely, positive pixels) being correctly segmented, where the click and visual information are fully fused by using a progressive three-tier interaction strategy. On the other hand, a general loss, Balanced Normalized Focal Loss (BNFL), is proposed. Its core is to use a group of control coefficients related to sample gradients and forces the network to pay more attention to positive and hard-to-segment pixels during training. As a result, BNFL always tends to obtain a balanced cut of positive and negative samples in the decision space. Theoretical analysis shows that the commonly used Focal and BCE losses can be regarded as special cases of BNFL. Experiment results of five well-recognized datasets have shown the superiority of the proposed CCF-BC method compared to other state-of-the-art methods. The source code is publicly available at https://github.com/lab206/CCF-BC.

Integration of diverse visual prompts like clicks, scribbles, and boxes in interactive image segmentation significantly facilitates users' interaction as well as improves interaction efficiency. However, existing studies primarily encode the position or pixel regions of prompts without considering the contextual areas around them, resulting in insufficient prompt feedback, which is not conducive to performance acceleration. To tackle this problem, this paper proposes a simple yet effective Probabilistic Visual Prompt Unified Transformer (PVPUFormer) for interactive image segmentation, which allows users to flexibly input diverse visual prompts with the probabilistic prompt encoding and feature post-processing to excavate sufficient and robust prompt features for performance boosting. Specifically, we first propose a Probabilistic Prompt-unified Encoder (PPuE) to generate a unified one-dimensional vector by exploring both prompt and non-prompt contextual information, offering richer feedback cues to accelerate performance improvement. On this basis, we further present a Prompt-to-Pixel Contrastive (P2C) loss to accurately align both prompt and pixel features, bridging the representation gap between them to offer consistent feature representations for mask prediction. Moreover, our approach designs a Dual-cross Merging Attention (DMA) module to implement bidirectional feature interaction between image and prompt features, generating notable features for performance improvement. A comprehensive variety of experiments on several challenging datasets demonstrates that the proposed components achieve consistent improvements, yielding state-of-the-art interactive segmentation performance. Our code is available at https://github.com/XuZhang1211/PVPUFormer.

Temporal information plays a pivotal role in Bird’s-Eye-View (BEV) driving scene understanding, which can alleviate the visual information sparsity. However, the indiscriminate temporal fusion method will cause the barrier of feature redundancy when constructing vectorized High-Definition (HD) maps. In this paper, we revisit the temporal fusion of vectorized HD maps, focusing on temporal instance consistency and temporal map consistency learning. To improve the representation of instances in single-frame maps, we introduce a novel method, DTCLMapper. This approach uses a dual-stream temporal consistency learning module that combines instance embedding with geometry maps. In the instance embedding component, our approach integrates temporal Instance Consistency Learning (ICL), ensuring consistency from vector points and instance features aggregated from points. A vectorized points pre-selection module is employed to enhance the regression efficiency of vector points from each instance. Then aggregated instance features obtained from the vectorized points preselection module are grounded in contrastive learning to realize temporal consistency, where positive and negative samples are selected based on position and semantic information. The geometry mapping component introduces Map Consistency Learning (MCL) designed with self-supervised learning. The MCL enhances the generalization capability of our consistent learning approach by concentrating on the global location and distribution constraints of the instances. Extensive experiments on well-recognized benchmarks indicate that the proposed DTCLMapper achieves state-of-the-art performance in vectorized mapping tasks, reaching 61.9% and 65.1% mAP scores on the nuScenes and Argoverse datasets, respectively. The source code is available athttps://github.com/lynn-yu/DTCLMapper.

This paper introduces the task of Auditory Referring Multi-Object Tracking (AR-MOT), which dynamically tracks specific objects in a video sequence based on audio expressions and appears as a challenging problem in autonomous driving. Due to the lack of semantic modeling capacity in audio and video, existing works have mainly focused on text-based multi-object tracking, which often comes at the cost of tracking quality, interaction efficiency, and even the safety of assistance systems, limiting the application of such methods in autonomous driving. In this paper, we delve into the problem of AR-MOT from the perspective of audio-video fusion and audio-video tracking. We put forward EchoTrack, an end-to-end AR-MOT framework with dual-stream vision transformers. The dual streams are intertwined with our Bidirectional Frequency-domain Cross-attention Fusion Module (Bi-FCFM), which bidirectionally fuses audio and video features from both frequency- and spatiotemporal domains. Moreover, we propose the Audio-visual Contrastive Tracking Learning (ACTL) regime to extract homogeneous semantic features between expressions and visual objects by learning homogeneous features between different audio and video objects effectively. Aside from the architectural design, we establish the first set of large-scale AR-MOT benchmarks, including Echo-KITTI, Echo-KITTI+, and Echo-BDD. Extensive experiments on the established benchmarks demonstrate the effectiveness of the proposed EchoTrack and its components. The source code and datasets are available athttps://github.com/lab206/EchoTrack.

Precisely predicting the drug-drug interaction (DDI) is an important application and host research topic in drug discovery, especially for avoiding the adverse effect when using drug combination treatment for patients. Nowadays, machine learning and deep learning methods have achieved great success in DDI prediction. However, we notice that most of the works ignore the importance of the relation type when building the DDI prediction models. In this work, we propose a novel R$^2$-DDI framework, which introduces a relation-aware feature refinement module for drug representation learning. The relation feature is integrated into drug representation and refined in the framework. With the refinement features, we also incorporate the consistency training method to regularize the multi-branch predictions for better generalization. Through extensive experiments and studies, we demonstrate our R$^2$-DDI approach can significantly improve the DDI prediction performance over multiple real-world datasets and settings, and our method shows better generalization ability with the help of the feature refinement design.

Point-based 3D detection approaches usually suffer from the severe point sampling imbalance problem between foreground and background. We observe that prior works have attempted to alleviate this imbalance by emphasizing foreground sampling. However, even adequate foreground sampling may be extremely unbalanced between nearby and distant objects, yielding unsatisfactory performance in detecting distant objects. To tackle this issue, this paper first proposes a novel method named Distant Object Augmented Set Abstraction and Regression (DO-SA&R) to enhance distant object detection, which is vital for the timely response of decision-making systems like autonomous driving. Technically, our approach first designs DO-SA with novel distant object augmented farthest point sampling (DO-FPS) to emphasize sampling on distant objects by leveraging both object-dependent and depth-dependent information. Then, we propose distant object augmented regression to reweight all the instance boxes for strengthening regression training on distant objects. In practice, the proposed DO-SA&R can be easily embedded into the existing modules, yielding consistent performance improvements, especially on detecting distant objects. Extensive experiments are conducted on the popular KITTI, nuScenes and Waymo datasets, and DO-SA&R demonstrates superior performance, especially for distant object detection. Our code is available at https://github.com/mikasa3lili/DO-SAR.