Yan Wang

Shuhan Tan, Kashyap Chitta, Yuxiao Chen et al.

Recent Vision-Language-Action (VLA) models for autonomous driving explore inference-time reasoning as a way to improve driving performance and safety in challenging scenarios. Most prior work uses natural language to express chain-of-thought (CoT) reasoning before producing driving actions. However, text may not be the most efficient representation for reasoning. In this work, we present Latent-CoT-Drive (LCDrive): a model that expresses CoT in a latent language that captures possible outcomes of the driving actions being considered. Our approach unifies CoT reasoning and decision making by representing both in an action-aligned latent space. Instead of natural language, the model reasons by interleaving (1) action-proposal tokens, which use the same vocabulary as the model's output actions; and (2) world model tokens, which are grounded in a learned latent world model and express future outcomes of these actions. We cold start latent CoT by supervising the model's action proposals and world model tokens based on ground-truth future rollouts of the scene. We then post-train with closed-loop reinforcement learning to strengthen reasoning capabilities. On a large-scale end-to-end driving benchmark, LCDrive achieves faster inference, better trajectory quality, and larger improvements from interactive reinforcement learning compared to both non-reasoning and text-reasoning baselines.

Jiageng Mao, Boyi Li, Boris Ivanovic et al.

Synthesizing photo-realistic visual observations from an ego vehicle's driving trajectory is a critical step towards scalable training of self-driving models. Reconstruction-based methods create 3D scenes from driving logs and synthesize geometry-consistent driving videos through neural rendering, but their dependence on costly object annotations limits their ability to generalize to in-the-wild driving scenarios. On the other hand, generative models can synthesize action-conditioned driving videos in a more generalizable way but often struggle with maintaining 3D visual consistency. In this paper, we present DreamDrive, a 4D spatial-temporal scene generation approach that combines the merits of generation and reconstruction, to synthesize generalizable 4D driving scenes and dynamic driving videos with 3D consistency. Specifically, we leverage the generative power of video diffusion models to synthesize a sequence of visual references and further elevate them to 4D with a novel hybrid Gaussian representation. Given a driving trajectory, we then render 3D-consistent driving videos via Gaussian splatting. The use of generative priors allows our method to produce high-quality 4D scenes from in-the-wild driving data, while neural rendering ensures 3D-consistent video generation from the 4D scenes. Extensive experiments on nuScenes and street view images demonstrate that DreamDrive can generate controllable and generalizable 4D driving scenes, synthesize novel views of driving videos with high fidelity and 3D consistency, decompose static and dynamic elements in a self-supervised manner, and enhance perception and planning tasks for autonomous driving.

Boris Ivanovic, Cristiano Saltori, Yurong You et al.

Autoregressive Transformers are increasingly being deployed as end-to-end robot and autonomous vehicle (AV) policy architectures, owing to their scalability and potential to leverage internet-scale pretraining for generalization. Accordingly, tokenizing sensor data efficiently is paramount to ensuring the real-time feasibility of such architectures on embedded hardware. To this end, we present an efficient triplane-based multi-camera tokenization strategy that leverages recent advances in 3D neural reconstruction and rendering to produce sensor tokens that are agnostic to the number of input cameras and their resolution, while explicitly accounting for their geometry around an AV. Experiments on a large-scale AV dataset and state-of-the-art neural simulator demonstrate that our approach yields significant savings over current image patch-based tokenization strategies, producing up to 72% fewer tokens, resulting in up to 50% faster policy inference while achieving the same open-loop motion planning accuracy and improved offroad rates in closed-loop driving simulations.

Jang Hyun Cho, Boris Ivanovic, Yulong Cao et al.

Multi-modal large language models (MLLMs) have shown incredible capabilities in a variety of 2D vision and language tasks. We extend MLLMs' perceptual capabilities to ground and reason about images in 3-dimensional space. To that end, we first develop a large-scale pre-training dataset for 2D and 3D called LV3D by combining multiple existing 2D and 3D recognition datasets under a common task formulation: as multi-turn question-answering. Next, we introduce a new MLLM named Cube-LLM and pre-train it on LV3D. We show that pure data scaling makes a strong 3D perception capability without 3D specific architectural design or training objective. Cube-LLM exhibits intriguing properties similar to LLMs: (1) Cube-LLM can apply chain-of-thought prompting to improve 3D understanding from 2D context information. (2) Cube-LLM can follow complex and diverse instructions and adapt to versatile input and output formats. (3) Cube-LLM can be visually prompted such as 2D box or a set of candidate 3D boxes from specialists. Our experiments on outdoor benchmarks demonstrate that Cube-LLM significantly outperforms existing baselines by 21.3 points of AP-BEV on the Talk2Car dataset for 3D grounded reasoning and 17.7 points on the DriveLM dataset for complex reasoning about driving scenarios, respectively. Cube-LLM also shows competitive results in general MLLM benchmarks such as refCOCO for 2D grounding with (87.0) average score, as well as visual question answering benchmarks such as VQAv2, GQA, SQA, POPE, etc. for complex reasoning. Our project is available at https://janghyuncho.github.io/Cube-LLM.

Tian Lan, Deng Cai, Yan Wang et al.

Recent progress in deep learning has continuously improved the accuracy of dialogue response selection. In particular, sophisticated neural network architectures are leveraged to capture the rich interactions between dialogue context and response candidates. While remarkably effective, these models also bring in a steep increase in computational cost. Consequently, such models can only be used as a re-rank module in practice. In this study, we present a solution to directly select proper responses from a large corpus or even a nonparallel corpus that only consists of unpaired sentences, using a dense retrieval model. To push the limits of dense retrieval, we design an interaction layer upon the dense retrieval models and apply a set of tailor-designed learning strategies. Our model shows superiority over strong baselines on the conventional re-rank evaluation setting, which is remarkable given its efficiency. To verify the effectiveness of our approach in realistic scenarios, we also conduct full-rank evaluation, where the target is to select proper responses from a full candidate pool that may contain millions of candidates and evaluate them fairly through human annotations. Our proposed model notably outperforms pipeline baselines that integrate fast recall and expressive re-rank modules. Human evaluation results show that enlarging the candidate pool with nonparallel corpora improves response quality further.

Yan Wang, Yuhang Li, Ruihao Gong

Systematic error, which is not determined by chance, often refers to the inaccuracy (involving either the observation or measurement process) inherent to a system. In this paper, we exhibit some long-neglected but frequent-happening adversarial examples caused by systematic error. More specifically, we find the trained neural network classifier can be fooled by inconsistent implementations of image decoding and resize. This tiny difference between these implementations often causes an accuracy drop from training to deployment. To benchmark these real-world adversarial examples, we propose ImageNet-S dataset, which enables researchers to measure a classifier's robustness to systematic error. For example, we find a normal ResNet-50 trained on ImageNet can have 1%-5% accuracy difference due to the systematic error. Together our evaluation and dataset may aid future work toward real-world robustness and practical generalization.

Lingyun Feng, Jianwei Yu, Deng Cai et al.

Language understanding in speech-based systems have attracted much attention in recent years with the growing demand for voice interface applications. However, the robustness of natural language understanding (NLU) systems to errors introduced by automatic speech recognition (ASR) is under-examined. %To facilitate the research on ASR-robust general language understanding, In this paper, we propose ASR-GLUE benchmark, a new collection of 6 different NLU tasks for evaluating the performance of models under ASR error across 3 different levels of background noise and 6 speakers with various voice characteristics. Based on the proposed benchmark, we systematically investigate the effect of ASR error on NLU tasks in terms of noise intensity, error type and speaker variants. We further purpose two ways, correction-based method and data augmentation-based method to improve robustness of the NLU systems. Extensive experimental results and analysises show that the proposed methods are effective to some extent, but still far from human performance, demonstrating that NLU under ASR error is still very challenging and requires further research.

Zelong Yang, Yan Wang, Piji Li et al.

The multiplayer online battle arena (MOBA) games have become increasingly popular in recent years. Consequently, many efforts have been devoted to providing pre-game or in-game predictions for them. However, these works are limited in the following two aspects: 1) the lack of sufficient in-game features; 2) the absence of interpretability in the prediction results. These two limitations greatly restrict the practical performance and industrial application of the current works. In this work, we collect and release a large-scale dataset containing rich in-game features for the popular MOBA game Honor of Kings. We then propose to predict four types of important events in an interpretable way by attributing the predictions to the input features using two gradient-based attribution methods: Integrated Gradients and SmoothGrad. To evaluate the explanatory power of different models and attribution methods, a fidelity-based evaluation metric is further proposed. Finally, we evaluate the accuracy and Fidelity of several competitive methods on the collected dataset to assess how well machines predict events in MOBA games.

Zelong Yang, Zhufeng Pan, Yan Wang et al.

With the rapid prevalence and explosive development of MOBA esports (Multiplayer Online Battle Arena electronic sports), much research effort has been devoted to automatically predicting game results (win predictions). While this task has great potential in various applications, such as esports live streaming and game commentator AI systems, previous studies fail to investigate the methods to interpret these win predictions. To mitigate this issue, we collected a large-scale dataset that contains real-time game records with rich input features of the popular MOBA game Honor of Kings. For interpretable predictions, we proposed a Two-Stage Spatial-Temporal Network (TSSTN) that can not only provide accurate real-time win predictions but also attribute the ultimate prediction results to the contributions of different features for interpretability. Experiment results and applications in real-world live streaming scenarios showed that the proposed TSSTN model is effective both in prediction accuracy and interpretability.