Consistency distillation methods have demonstrated significant success in accelerating generative tasks of diffusion models. However, since previous consistency distillation methods use simple and straightforward strategies in selecting target timesteps, they usually struggle with blurs and detail losses in generated images. To address these limitations, we introduce Target-Driven Distillation (TDD), which (1) adopts a delicate selection strategy of target timesteps, increasing the training efficiency; (2) utilizes decoupled guidances during training, making TDD open to post-tuning on guidance scale during inference periods; (3) can be optionally equipped with non-equidistant sampling and x0 clipping, enabling a more flexible and accurate way for image sampling. Experiments verify that TDD achieves state-of-the-art performance in few-step generation, offering a better choice among consistency distillation models.

Diffusion models have demonstrated remarkable capabilities in generating high-quality images. Recent advancements in Layout-to-Image (L2I) generation have leveraged positional conditions and textual descriptions to facilitate precise and controllable image synthesis. Despite overall progress, current L2I methods still exhibit suboptimal performance. Therefore, we propose InstanceAssemble, a novel architecture that incorporates layout conditions via instance-assembling attention, enabling position control with bounding boxes (bbox) and multimodal content control including texts and additional visual content. Our method achieves flexible adaption to existing DiT-based T2I models through light-weighted LoRA modules. Additionally, we propose a Layout-to-Image benchmark, Denselayout, a comprehensive benchmark for layout-to-image generation, containing 5k images with 90k instances in total. We further introduce Layout Grounding Score (LGS), an interpretable evaluation metric to more precisely assess the accuracy of L2I generation. Experiments demonstrate that our InstanceAssemble method achieves state-of-the-art performance under complex layout conditions, while exhibiting strong compatibility with diverse style LoRA modules. The code and pretrained models are publicly available at \url{https://github.com/FireRedTeam/InstanceAssemble}.

Recent advancements in subject-driven image generation have led to zero-shot generation, yet precise selection and focus on crucial subject representations remain challenging. Addressing this, we introduce the SSR-Encoder, a novel architecture designed for selectively capturing any subject from single or multiple reference images. It responds to various query modalities including text and masks, without necessitating test-time fine-tuning. The SSR-Encoder combines a Token-to-Patch Aligner that aligns query inputs with image patches and a Detail-Preserving Subject Encoder for extracting and preserving fine features of the subjects, thereby generating subject embeddings. These embeddings, used in conjunction with original text embeddings, condition the generation process. Characterized by its model generalizability and efficiency, the SSR-Encoder adapts to a range of custom models and control modules. Enhanced by the Embedding Consistency Regularization Loss for improved training, our extensive experiments demonstrate its effectiveness in versatile and high-quality image generation, indicating its broad applicability. Project page: ssr-encoder.github.io

Point cloud completion aims to recover accurate global geometry and preserve fine-grained local details from partial point clouds. Conventional methods typically predict unseen points directly from 3D point cloud coordinates or use self-projected multi-view depth maps to ease this task. However, these gray-scale depth maps cannot reach multi-view consistency, consequently restricting the performance. In this paper, we introduce a GeoFormer that simultaneously enhances the global geometric structure of the points and improves the local details. Specifically, we design a CCM Feature Enhanced Point Generator to integrate image features from multi-view consistent canonical coordinate maps (CCMs) and align them with pure point features, thereby enhancing the global geometry feature. Additionally, we employ the Multi-scale Geometry-aware Upsampler module to progressively enhance local details. This is achieved through cross attention between the multi-scale features extracted from the partial input and the features derived from previously estimated points. Extensive experiments on the PCN, ShapeNet-55/34, and KITTI benchmarks demonstrate that our GeoFormer outperforms recent methods, achieving the state-of-the-art performance. Our code is available at https://github.com/Jinpeng-Yu/GeoFormer.

Diverse human motion prediction (HMP) is a fundamental application in computer vision that has recently attracted considerable interest. Prior methods primarily focus on the stochastic nature of human motion, while neglecting the specific impact of external environment, leading to the pronounced artifacts in prediction when applied to real-world scenarios. To fill this gap, this work introduces a novel task: predicting diverse human motion within real-world 3D scenes. In contrast to prior works, it requires harmonizing the deterministic constraints imposed by the surrounding 3D scenes with the stochastic aspect of human motion. For this purpose, we propose DiMoP3D, a diverse motion prediction framework with 3D scene awareness, which leverages the 3D point cloud and observed sequence to generate diverse and high-fidelity predictions. DiMoP3D is able to comprehend the 3D scene, and determines the probable target objects and their desired interactive pose based on the historical motion. Then, it plans the obstacle-free trajectory towards these interested objects, and generates diverse and physically-consistent future motions. On top of that, DiMoP3D identifies deterministic factors in the scene and integrates them into the stochastic modeling, making the diverse HMP in realistic scenes become a controllable stochastic generation process. On two real-captured benchmarks, DiMoP3D has demonstrated significant improvements over state-of-the-art methods, showcasing its effectiveness in generating diverse and physically-consistent motion predictions within real-world 3D environments.

Prior multi-frame optical flow methods typically estimate flow repeatedly in a pair-wise manner, leading to significant computational redundancy. To mitigate this, we implement a Streamlined In-batch Multi-frame (SIM) pipeline, specifically tailored to video inputs to minimize redundant calculations. It enables the simultaneous prediction of successive unidirectional flows in a single forward pass, boosting processing speed by 44.43% and reaching efficiencies on par with two-frame networks. Moreover, we investigate various spatiotemporal modeling methods for optical flow estimation within this pipeline. Notably, we propose a simple yet highly effective parameter-efficient Integrative spatiotemporal Coherence (ISC) modeling method, alongside a lightweight Global Temporal Regressor (GTR) to harness temporal cues. The proposed ISC and GTR bring powerful spatiotemporal modeling capabilities and significantly enhance accuracy, including in occluded areas, while adding modest computations to the SIM pipeline. Compared to the baseline, our approach, StreamFlow, achieves performance enhancements of 15.45% and 11.37% on the Sintel clean and final test sets respectively, with gains of 15.53% and 10.77% on occluded regions and only a 1.11% rise in latency. Furthermore, StreamFlow exhibits state-of-the-art cross-dataset testing results on Sintel and KITTI, demonstrating its robust cross-domain generalization capabilities. The code is available [here](https://github.com/littlespray/StreamFlow).

Recently, deep learning-based facial landmark detection has achieved significant improvement. However, the semantic ambiguity problem degrades detection performance. Specifically, the semantic ambiguity causes inconsistent annotation and negatively affects the model's convergence, leading to worse accuracy and instability prediction. To solve this problem, we propose a Self-adapTive Ambiguity Reduction (STAR) loss by exploiting the properties of se-mantic ambiguity. We find that semantic ambiguity results in the anisotropic predicted distribution, which inspires us to use predicted distribution to represent semantic ambiguity. Based on this, we design the STAR loss that measures the anisotropism of the predicted distribution. Compared with the standard regression loss, STAR loss is encouraged to be small when the predicted distribution is anisotropic and thus adaptively mitigates the impact of semantic ambiguity. Moreover, we propose two kinds of eigen-value restriction methods that could avoid both distribution's abnormal change and the model's premature convergence. Finally, the comprehensive experiments demonstrate that STAR loss outperforms the state-of-the-art methods on three benchmarks, i.e., COFW, 300W, and WFLW, with negligible computation overhead. Code is at https://github.com/ZhenglinZhou/STAR

Industrial automated production systems are mechatronic and long living systems that undergo changing requirements throughout their life cycle. While the proportion of functionality implemented by software is growing, adjustments are usually implemented using a clone-and-own principle, which results in unmanaged software variants and versions. Furthermore, the need for adapting the control software also results from changes in other disciplines such as mechanical or electrical/electronics. The various drawbacks on software maintainability, that are provoked through clone-andown, call for a shift to modular development. As a first step to realize this migration, software projects need to be analyzed in terms of variability. Secondly, visualization patterns reflecting variability are needed to present the analysis results to domain experts. However, choosing an appropriate visualization is challenging as different domain experts pursue different aims, which should be supported by the visualization and might even require different levels of detail. In this paper, visualization patterns for three different use scenarios are proposed and evaluated using an apprentice group and industrial expert feedback.

Automated production systems become more and more complex. This makes it increasingly difficult to keep track of performed changes and already executed test cases. This endangers the systems quality as the risk of missing important test cases while planning the test execution is high, especially for testers with little experience. To face this challenge, testers should be supported by an automatic test prioritisation based on metrics in selecting the right test cases for the test execution. In industry, many different test prioritisation criteria and strategies are used for this purpose. In an industrial interview, experts discussed and ranked prioritisation criteria that are currently used within the respective companies. As a result, this paper presents the cactus prioritisation model, which graphically resembles the industrial ranking and weighting of the criteria. Based on the prioritisation cactus and its criteria, a simple prioritisation metric is introduced to determine the utility of each test case regarding the system under test. The test cases are prioritised according to their descending utility. Furthermore, approaches and metrics to realise the different individual prioritisation criteria are proposed.

Due to the increasing integration of different disciplines, Model-based Systems Engineering (MBSE) is becoming more beneficial for the development of mechatronic systems. However, even though all relevant discipline-specific models can be integrated into one system model, how to maintain consistency between a discipline-specific model and the central system model during model updates still needs to be considered. In this paper, an approach for automatic synchronization of a discipline-specific model i.e. mechanical CAD model and the central SysML-based model i.e. a SysML4 Mechatronics plant model, is introduced. AutomationML(AML), a data exchange format which enables a lossless exchange of information between different discipline-specific engineering tools, is applied during the model synchronization. Mapping rules are proposed to facilitate a consistent round-trip engineering and transformation. Feasibility and applicability of this approach are demonstrated through a use case on a production plant with two applied commercial tools, PTC Integrity Modeler and Creo Parametric.

Increasing integration of different disciplines such as mechanics, electrics/electronics and control software has led to higher complexity in the development of an automated production system (aPS), bringing forth some challenges along with it. In order to address these challenges, a multi-disciplinary engineering workflow can be built using a model-based system engineering (MBSE) approach. However, the various tools in different disciplines impede the efficient collaboration of different models. Because of the implicit dependencies between different disciplines, it is difficult to ensure the consistency after a change. This paper provides an approach to building an efficient multi-disciplinary engineering workflow based on MBSE. Through a combination of SysML4Mechatronics modeling language with inconsistency management and Product Lifecycle Management (PLM), the proposed engineering workflow facilitates the design of multi-disciplinary aPS and supports information tracking, consistency assurance, as well as change analysis across disciplines. The benefits of the proposed approach are demonstrated with a case study on a lab-sized production plant.