Kaiqi Liu

Jiawei Han, Kaiqi Liu, Wei Li et al.

Point cloud semantic segmentation can significantly enhance the perception of an intelligent agent. Nevertheless, the discriminative capability of the segmentation network is influenced by the quantity of samples available for different categories. To mitigate the cognitive bias induced by class imbalance, this paper introduces a novel method, namely subspace prototype guidance (\textbf{SPG}), to guide the training of segmentation network. Specifically, the point cloud is initially separated into independent point sets by category to provide initial conditions for the generation of feature subspaces. The auxiliary branch which consists of an encoder and a projection head maps these point sets into separate feature subspaces. Subsequently, the feature prototypes which are extracted from the current separate subspaces and then combined with prototypes of historical subspaces guide the feature space of main branch to enhance the discriminability of features of minority categories. The prototypes derived from the feature space of main branch are also employed to guide the training of the auxiliary branch, forming a supervisory loop to maintain consistent convergence of the entire network. The experiments conducted on the large public benchmarks (i.e. S3DIS, ScanNet v2, ScanNet200, Toronto-3D) and collected real-world data illustrate that the proposed method significantly improves the segmentation performance and surpasses the state-of-the-art method. The code is available at \url{https://github.com/Javion11/PointLiBR.git}.

Ziqi Cai, Shuchen Weng, Kaiqi Liu et al.

Recovering dynamic environment maps from a single in-the-wild video is crucial for photorealistic rendering, yet remains a challenge. Recent video generation models can produce photorealistic scenes with complex lighting, possessing an inherent understanding of lighting. In this paper, we introduce V-LITE (Video generation models are inherent lighting estimators), a framework that unlocks this internal knowledge by reframing lighting estimation as a guided video inpainting task. Inspired by VFX industry practices, we insert a synthetic chrome ball into the scene to compel the model to generate physically plausible reflections from the surrounding spatio-temporal context. To bridge the gap from LDR-native models to the HDR domain, we design an HDR-aware VAE and employ an efficient LoRA-based fine-tuning strategy. We then construct a mixed dataset comprising high-fidelity HDR images to provide realistic HDR priors, and in-the-wild HDR videos to provide dynamic spatio-temporal context. Extensive experiments demonstrate that V-LITE produces temporally coherent HDR environment maps, revealing that modern video diffusion models are not merely synthesizers but also powerful, inherently capable estimators of physical scene lighting.

Kaiqi Liu, Yunyao Mao, Ziqi Cai et al.

While recent generative models produce high-fidelity videos, they struggle with the complex narrative control required for coherent multi-shot audio-visual generation. Existing methods suffer from temporal misalignment, limited controllability, and incomplete scripting. In this paper, we propose MAVIN, the first framework for multi-shot audio-visual generation with customized narrative control. To resolve temporal misalignment, we propose boundary-aware attention, which leverages hierarchical captions and boundary-aware token routing to render audio-visual elements within their respective temporal boundaries. To improve the controllability for multi-subject scenarios, we propose ID-aware propagation, utilizing identity embeddings and an identity-aware mask to bind specific identities to consistent visual appearances and vocal timbres. To provide comprehensive audio-visual narratives, we present a multi-agent scripting pipeline to transform free-form user inputs into hierarchical captions. Furthermore, we construct MAVINSet, a multi-shot audio-visual dataset for robust training and evaluation. Extensive experiments demonstrate that MAVIN achieves state-of-the-art performance, opening up a new avenue for integrating generative models into professional filmmaking workflows.

Jing Wang, Xiangxin Zhou, Jiajun Liang et al.

Autoregressive (AR) video diffusion models enable low-latency streaming generation by synthesizing videos chunk by chunk with cached visual context, but this chunk-wise formulation makes temporal instruction following ambiguous. A single global prompt does not specify which sub-event should be realized in each chunk, while naively switching to step-wise prompts often leads to delayed reactions, blended step semantics, and error propagation across prompt transitions. These failures are difficult to address with supervised fine-tuning or distillation alone: SFT suffers from exposure bias, while rollout-based distillation still optimizes low-level denoising or teacher-distribution matching rather than directly enforcing action ordering and prompt-transition correctness. We address these challenges with TempAct, a planner--executor reinforcement learning framework that jointly optimizes temporal decomposition and step-conditioned execution for temporally plausible AR video generation. TempAct uses an LLM planner to explore span-aware step prompts that are executable by the video model, and trains an AR diffusion executor to follow these prompts under its own generated histories. Its key mechanism is hierarchical group exploration: candidate plans form planning groups, and each plan induces an execution group of multiple continuations from a shared visual context, enabling plan-level credit assignment for long-horizon temporal outcomes and executor-level credit assignment for prompt-switch behavior. We further design hierarchical rewards that combine plan-quality and full-video temporal feedback for the planner with local transition-level step-following rewards, aesthetic regularization, and KL constraints for the executor. Experiments on Self-Forcing and LongLive show that TempAct improves temporal consistency while preserving overall visual quality.

Jinghan Yu, Junhao Xiao, Zhiyuan Ma et al.

Recent years have witnessed the success of diffusion models in image customization tasks. However, existing mask-guided human erasing methods still struggle in complex scenarios such as human-human occlusion, human-object entanglement, and human-background interference, mainly due to the lack of large-scale multi-instance datasets and effective spatial decoupling to separate foreground from background. To bridge these gaps, we curate the MILD dataset capturing diverse poses, occlusions, and complex multi-instance interactions. We then define the Cross-Domain Attention Gap (CAG), an attention-gap metric to quantify semantic leakage. On top of these, we propose Multi-Layer Diffusion (MILD), which decomposes the generation process into independent denoising pathways, enabling separate reconstruction of each foreground instance and the background. To enhance human-centric understanding, we introduce Human Morphology Guidance, a plug-and-play module that incorporates pose, parsing, and spatial relationships into the diffusion process to improve structural awareness and restoration quality. Additionally, we present Spatially-Modulated Attention, an adaptive mechanism that leverages spatial mask priors to modulate attention across semantic regions, further widening the CAG to effectively minimize boundary artifacts and mitigate semantic leakage. Experiments show that MILD significantly outperforms existing methods. Datasets and code are publicly available at: https://mild-multi-layer-diffusion.github.io/.

Dehao Wang, Haohang Zhu, Yiwen Xu et al.

Road potholes pose a serious threat to driving safety and comfort, making their detection and assessment a critical task in fields such as autonomous driving. When driving vehicles, the operators usually avoid large potholes and approach smaller ones at reduced speeds to ensure safety. Therefore, accurately estimating pothole area is of vital importance. Most existing vision-based methods rely on distance priors to construct geometric models. However, their performance is susceptible to variations in camera angles and typically relies on the assumption of a flat road surface, potentially leading to significant errors in complex real-world environments. To address these problems, a robust pothole area estimation framework that integrates object detection and monocular depth estimation in a video stream is proposed in this paper. First, to enhance pothole feature extraction and improve the detection of small potholes, ACSH-YOLOv8 is proposed with ACmix module and the small object detection head. Then, the BoT-SORT algorithm is utilized for pothole tracking, while DepthAnything V2 generates depth maps for each frame. With the obtained depth maps and potholes labels, a novel Minimum Bounding Triangulated Pixel (MBTP) method is proposed for pothole area estimation. Finally, Kalman Filter based on Confidence and Distance (CDKF) is developed to maintain consistency of estimation results across consecutive frames. The results show that ACSH-YOLOv8 model achieves an AP(50) of 76.6%, representing a 7.6% improvement over YOLOv8. Through CDKF optimization across consecutive frames, pothole predictions become more robust, thereby enhancing the method's practical applicability.