Jialiang Wang

Chenfeng Xu, Bichen Wu, Ji Hou et al.

We present NeRF-Det, a novel method for indoor 3D detection with posed RGB images as input. Unlike existing indoor 3D detection methods that struggle to model scene geometry, our method makes novel use of NeRF in an end-to-end manner to explicitly estimate 3D geometry, thereby improving 3D detection performance. Specifically, to avoid the significant extra latency associated with per-scene optimization of NeRF, we introduce sufficient geometry priors to enhance the generalizability of NeRF-MLP. Furthermore, we subtly connect the detection and NeRF branches through a shared MLP, enabling an efficient adaptation of NeRF to detection and yielding geometry-aware volumetric representations for 3D detection. Our method outperforms state-of-the-arts by 3.9 mAP and 3.1 mAP on the ScanNet and ARKITScenes benchmarks, respectively. We provide extensive analysis to shed light on how NeRF-Det works. As a result of our joint-training design, NeRF-Det is able to generalize well to unseen scenes for object detection, view synthesis, and depth estimation tasks without requiring per-scene optimization. Code is available at \url{https://github.com/facebookresearch/NeRF-Det}.

Xiaoliang Dai, Ji Hou, Chih-Yao Ma et al. · meta-ai

Training text-to-image models with web scale image-text pairs enables the generation of a wide range of visual concepts from text. However, these pre-trained models often face challenges when it comes to generating highly aesthetic images. This creates the need for aesthetic alignment post pre-training. In this paper, we propose quality-tuning to effectively guide a pre-trained model to exclusively generate highly visually appealing images, while maintaining generality across visual concepts. Our key insight is that supervised fine-tuning with a set of surprisingly small but extremely visually appealing images can significantly improve the generation quality. We pre-train a latent diffusion model on $1.1$ billion image-text pairs and fine-tune it with only a few thousand carefully selected high-quality images. The resulting model, Emu, achieves a win rate of $82.9\%$ compared with its pre-trained only counterpart. Compared to the state-of-the-art SDXLv1.0, Emu is preferred $68.4\%$ and $71.3\%$ of the time on visual appeal on the standard PartiPrompts and our Open User Input benchmark based on the real-world usage of text-to-image models. In addition, we show that quality-tuning is a generic approach that is also effective for other architectures, including pixel diffusion and masked generative transformer models.

Jialiang Wang, Daniel Scharstein, Akash Bapat et al.

We present the design of a productionized end-to-end stereo depth sensing system that does pre-processing, online stereo rectification, and stereo depth estimation with a fallback to monocular depth estimation when rectification is unreliable. The output of our depth sensing system is then used in a novel view generation pipeline to create 3D computational photography effects using point-of-view images captured by smart glasses. All these steps are executed on-device on the stringent compute budget of a mobile phone, and because we expect the users can use a wide range of smartphones, our design needs to be general and cannot be dependent on a particular hardware or ML accelerator such as a smartphone GPU. Although each of these steps is well studied, a description of a practical system is still lacking. For such a system, all these steps need to work in tandem with one another and fallback gracefully on failures within the system or less than ideal input data. We show how we handle unforeseen changes to calibration, e.g., due to heat, robustly support depth estimation in the wild, and still abide by the memory and latency constraints required for a smooth user experience. We show that our trained models are fast, and run in less than 1s on a six-year-old Samsung Galaxy S8 phone's CPU. Our models generalize well to unseen data and achieve good results on Middlebury and in-the-wild images captured from the smart glasses.

Jialiang Wang, Haotian Wei, Yi Wang et al.

Human activity recognition (HAR) based on multimodal sensors has become a rapidly growing branch of biometric recognition and artificial intelligence. However, how to fully mine multimodal time series data and effectively learn accurate behavioral features has always been a hot topic in this field. Practical applications also require a well-generalized framework that can quickly process a variety of raw sensor data and learn better feature representations. This paper proposes a universal multi-sensor network (UMSNet) for human activity recognition. In particular, we propose a new lightweight sensor residual block (called LSR block), which improves the performance by reducing the number of activation function and normalization layers, and adding inverted bottleneck structure and grouping convolution. Then, the Transformer is used to extract the relationship of series features to realize the classification and recognition of human activities. Our framework has a clear structure and can be directly applied to various types of multi-modal Time Series Classification (TSC) tasks after simple specialization. Extensive experiments show that the proposed UMSNet outperforms other state-of-the-art methods on two popular multi-sensor human activity recognition datasets (i.e. HHAR dataset and MHEALTH dataset).

Yuanhao Cai, Kunpeng Li, Menglin Jia et al.

Recent advances in text-to-video (T2V) generation have achieved good visual quality, yet synthesizing videos that faithfully follow physical laws remains an open challenge. Existing methods mainly based on graphics or prompt extension struggle to generalize beyond simple simulated environments or learn implicit physical reasoning. The scarcity of training data with rich physics interactions and phenomena is also a problem. In this paper, we first introduce a Physics-Augmented video data construction Pipeline, PhyAugPipe, that leverages a vision-language model (VLM) with chain-of-thought reasoning to collect a large-scale training dataset, PhyVidGen-135K. Then we formulate a principled Physics-aware Groupwise Direct Preference Optimization, PhyGDPO, framework that uses real-world video as winning case to guarantee correct physics learning and builds upon the groupwise Plackett-Luce probabilistic model to capture holistic preferences beyond pairwise comparisons. In PhyGDPO, we design a Physics-Guided Rewarding (PGR) scheme that leverages VLM-based physical rewards to direct the optimization to focus on challenging physics cases. In addition, we propose a LoRA-Switch Reference (LoRA-SR) scheme that avoids full-model duplication as reference for efficient DPO training. Experiments show that our method significantly outperforms state-of-the-art open-source methods on PhyGenBench and VideoPhy2. Please check our project page at https://caiyuanhao1998.github.io/project/PhyGDPO for more video results. Our code, models, and data will be released at https://github.com/caiyuanhao1998/Open-PhyGDPO

Christina Zhang, Simran Motwani, Matthew Yu et al.

Latent diffusion models (LDMs) have made significant advancements in the field of image generation in recent years. One major advantage of LDMs is their ability to operate in a compressed latent space, allowing for more efficient training and deployment. However, despite these advantages, challenges with LDMs still remain. For example, it has been observed that LDMs often generate high-frequency details and complex compositions imperfectly. We hypothesize that one reason for these flaws is due to the fact that all pre- and post-training of LDMs are done in latent space, which is typically $8 \times 8$ lower spatial-resolution than the output images. To address this issue, we propose adding pixel-space supervision in the post-training process to better preserve high-frequency details. Experimentally, we show that adding a pixel-space objective significantly improves both supervised quality fine-tuning and preference-based post-training by a large margin on a state-of-the-art DiT transformer and U-Net diffusion models in both visual quality and visual flaw metrics, while maintaining the same text alignment quality.

Kunpeng Song, Tingbo Hou, Zecheng He et al.

In this paper, we introduce DirectorLLM, a novel video generation model that employs a large language model (LLM) to orchestrate human poses within videos. As foundational text-to-video models rapidly evolve, the demand for high-quality human motion and interaction grows. To address this need and enhance the authenticity of human motions, we extend the LLM from a text generator to a video director and human motion simulator. Utilizing open-source resources from Llama 3, we train the DirectorLLM to generate detailed instructional signals, such as human poses, to guide video generation. This approach offloads the simulation of human motion from the video generator to the LLM, effectively creating informative outlines for human-centric scenes. These signals are used as conditions by the video renderer, facilitating more realistic and prompt-following video generation. As an independent LLM module, it can be applied to different video renderers, including UNet and DiT, with minimal effort. Experiments on automatic evaluation benchmarks and human evaluations show that our model outperforms existing ones in generating videos with higher human motion fidelity, improved prompt faithfulness, and enhanced rendered subject naturalness.

Jialiang Wang, Yurong Zhong, Weiling Li

Latent Factor (LF) models are effective in representing high-dimension and sparse (HiDS) data via low-rank matrices approximation. Hessian-free (HF) optimization is an efficient method to utilizing second-order information of an LF model's objective function and it has been utilized to optimize second-order LF (SLF) model. However, the low-rank representation ability of a SLF model heavily relies on its multiple hyperparameters. Determining these hyperparameters is time-consuming and it largely reduces the practicability of an SLF model. To address this issue, a practical SLF (PSLF) model is proposed in this work. It realizes hyperparameter self-adaptation with a distributed particle swarm optimizer (DPSO), which is gradient-free and parallelized. Experiments on real HiDS data sets indicate that PSLF model has a competitive advantage over state-of-the-art models in data representation ability.

Jiahao Zhou, Chen Long, Yue Xie et al.

End-to-end models capable of handling multiple sub-tasks in parallel have become a new trend, thereby presenting significant challenges and opportunities for the integration of multiple tasks within the domain of 3D vision. The limitations of 3D data acquisition conditions have not only restricted the exploration of many innovative research problems but have also caused existing 3D datasets to predominantly focus on single tasks. This has resulted in a lack of systematic approaches and theoretical frameworks for 3D multi-task learning, with most efforts merely serving as auxiliary support to the primary task. In this paper, we introduce WHU-Synthetic, a large-scale 3D synthetic perception dataset designed for multi-task learning, from the initial data augmentation (upsampling and depth completion), through scene understanding (segmentation), to macro-level tasks (place recognition and 3D reconstruction). Collected in the same environmental domain, we ensure inherent alignment across sub-tasks to construct multi-task models without separate training methods. Besides, we implement several novel settings, making it possible to realize certain ideas that are difficult to achieve in real-world scenarios. This supports more adaptive and robust multi-task perception tasks, such as sampling on city-level models, providing point clouds with different densities, and simulating temporal changes. Using our dataset, we conduct several experiments to investigate mutual benefits between sub-tasks, revealing new observations, challenges, and opportunities for future research. The dataset is accessible at https://github.com/WHU-USI3DV/WHU-Synthetic.

Yiyuan Zhu, Yongjun Li, Jialiang Wang et al.

Over the past years, a large number of fake news detection algorithms based on deep learning have emerged. However, they are often developed under different frameworks, each mandating distinct utilization methodologies, consequently hindering reproducibility. Additionally, a substantial amount of redundancy characterizes the code development of such fake news detection models. To address these concerns, we propose FaKnow, a unified and comprehensive fake news detection algorithm library. It encompasses a variety of widely used fake news detection models, categorized as content-based and social context-based approaches. This library covers the full spectrum of the model training and evaluation process, effectively organizing the data, models, and training procedures within a unified framework. Furthermore, it furnishes a series of auxiliary functionalities and tools, including visualization, and logging. Our work contributes to the standardization and unification of fake news detection research, concurrently facilitating the endeavors of researchers in this field. The open-source code and documentation can be accessed at https://github.com/NPURG/FaKnow and https://faknow.readthedocs.io, respectively.

Felix Wimbauer, Bichen Wu, Edgar Schoenfeld et al.

Diffusion models have recently revolutionized the field of image synthesis due to their ability to generate photorealistic images. However, one of the major drawbacks of diffusion models is that the image generation process is costly. A large image-to-image network has to be applied many times to iteratively refine an image from random noise. While many recent works propose techniques to reduce the number of required steps, they generally treat the underlying denoising network as a black box. In this work, we investigate the behavior of the layers within the network and find that 1) the layers' output changes smoothly over time, 2) the layers show distinct patterns of change, and 3) the change from step to step is often very small. We hypothesize that many layer computations in the denoising network are redundant. Leveraging this, we introduce block caching, in which we reuse outputs from layer blocks of previous steps to speed up inference. Furthermore, we propose a technique to automatically determine caching schedules based on each block's changes over timesteps. In our experiments, we show through FID, human evaluation and qualitative analysis that Block Caching allows to generate images with higher visual quality at the same computational cost. We demonstrate this for different state-of-the-art models (LDM and EMU) and solvers (DDIM and DPM).

Jialiang Wang, Weiling Li, Yurong Zhong et al.

Interactions among large number of entities is naturally high-dimensional and incomplete (HDI) in many big data related tasks. Behavioral characteristics of users are hidden in these interactions, hence, effective representation of the HDI data is a fundamental task for understanding user behaviors. Latent factor analysis (LFA) model has proven to be effective in representing HDI data. The performance of an LFA model relies heavily on its training process, which is a non-convex optimization. It has been proven that incorporating local curvature and preprocessing gradients during its training process can lead to superior performance compared to LFA models built with first-order family methods. However, with the escalation of data volume, the feasibility of second-order algorithms encounters challenges. To address this pivotal issue, this paper proposes a mini-block diagonal hessian-free (Mini-Hes) optimization for building an LFA model. It leverages the dominant diagonal blocks in the generalized Gauss-Newton matrix based on the analysis of the Hessian matrix of LFA model and serves as an intermediary strategy bridging the gap between first-order and second-order optimization methods. Experiment results indicate that, with Mini-Hes, the LFA model outperforms several state-of-the-art models in addressing missing data estimation task on multiple real HDI datasets from recommender system. (The source code of Mini-Hes is available at https://github.com/Goallow/Mini-Hes)

Feng Liang, Bichen Wu, Jialiang Wang et al.

Diffusion models have transformed the image-to-image (I2I) synthesis and are now permeating into videos. However, the advancement of video-to-video (V2V) synthesis has been hampered by the challenge of maintaining temporal consistency across video frames. This paper proposes a consistent V2V synthesis framework by jointly leveraging spatial conditions and temporal optical flow clues within the source video. Contrary to prior methods that strictly adhere to optical flow, our approach harnesses its benefits while handling the imperfection in flow estimation. We encode the optical flow via warping from the first frame and serve it as a supplementary reference in the diffusion model. This enables our model for video synthesis by editing the first frame with any prevalent I2I models and then propagating edits to successive frames. Our V2V model, FlowVid, demonstrates remarkable properties: (1) Flexibility: FlowVid works seamlessly with existing I2I models, facilitating various modifications, including stylization, object swaps, and local edits. (2) Efficiency: Generation of a 4-second video with 30 FPS and 512x512 resolution takes only 1.5 minutes, which is 3.1x, 7.2x, and 10.5x faster than CoDeF, Rerender, and TokenFlow, respectively. (3) High-quality: In user studies, our FlowVid is preferred 45.7% of the time, outperforming CoDeF (3.5%), Rerender (10.2%), and TokenFlow (40.4%).

Jonas Schult, Sam Tsai, Lukas Höllein et al.

Manually creating 3D environments for AR/VR applications is a complex process requiring expert knowledge in 3D modeling software. Pioneering works facilitate this process by generating room meshes conditioned on textual style descriptions. Yet, many of these automatically generated 3D meshes do not adhere to typical room layouts, compromising their plausibility, e.g., by placing several beds in one bedroom. To address these challenges, we present ControlRoom3D, a novel method to generate high-quality room meshes. Central to our approach is a user-defined 3D semantic proxy room that outlines a rough room layout based on semantic bounding boxes and a textual description of the overall room style. Our key insight is that when rendered to 2D, this 3D representation provides valuable geometric and semantic information to control powerful 2D models to generate 3D consistent textures and geometry that aligns well with the proxy room. Backed up by an extensive study including quantitative metrics and qualitative user evaluations, our method generates diverse and globally plausible 3D room meshes, thus empowering users to design 3D rooms effortlessly without specialized knowledge.

Hongjie Wang, Chih-Yao Ma, Yen-Cheng Liu et al.

Text-to-video generation enhances content creation but is highly computationally intensive: The computational cost of Diffusion Transformers (DiTs) scales quadratically in the number of pixels. This makes minute-length video generation extremely expensive, limiting most existing models to generating videos of only 10-20 seconds length. We propose a Linear-complexity text-to-video Generation (LinGen) framework whose cost scales linearly in the number of pixels. For the first time, LinGen enables high-resolution minute-length video generation on a single GPU without compromising quality. It replaces the computationally-dominant and quadratic-complexity block, self-attention, with a linear-complexity block called MATE, which consists of an MA-branch and a TE-branch. The MA-branch targets short-to-long-range correlations, combining a bidirectional Mamba2 block with our token rearrangement method, Rotary Major Scan, and our review tokens developed for long video generation. The TE-branch is a novel TEmporal Swin Attention block that focuses on temporal correlations between adjacent tokens and medium-range tokens. The MATE block addresses the adjacency preservation issue of Mamba and improves the consistency of generated videos significantly. Experimental results show that LinGen outperforms DiT (with a 75.6% win rate) in video quality with up to 15$\times$ (11.5$\times$) FLOPs (latency) reduction. Furthermore, both automatic metrics and human evaluation demonstrate our LinGen-4B yields comparable video quality to state-of-the-art models (with a 50.5%, 52.1%, 49.1% win rate with respect to Gen-3, LumaLabs, and Kling, respectively). This paves the way to hour-length movie generation and real-time interactive video generation. We provide 68s video generation results and more examples in our project website: https://lineargen.github.io/.

Yuewei Yang, Xiaoliang Dai, Jialiang Wang et al.

Latent Diffusion Models (LDMs) capture the dynamic evolution of latent variables over time, blending patterns and multimodality in a generative system. Despite the proficiency of LDM in various applications, such as text-to-image generation, facilitated by robust text encoders and a variational autoencoder, the critical need to deploy large generative models on edge devices compels a search for more compact yet effective alternatives. Post Training Quantization (PTQ), a method to compress the operational size of deep learning models, encounters challenges when applied to LDM due to temporal and structural complexities. This study proposes a quantization strategy that efficiently quantize LDMs, leveraging Signal-to-Quantization-Noise Ratio (SQNR) as a pivotal metric for evaluation. By treating the quantization discrepancy as relative noise and identifying sensitive part(s) of a model, we propose an efficient quantization approach encompassing both global and local strategies. The global quantization process mitigates relative quantization noise by initiating higher-precision quantization on sensitive blocks, while local treatments address specific challenges in quantization-sensitive and time-sensitive modules. The outcomes of our experiments reveal that the implementation of both global and local treatments yields a highly efficient and effective Post Training Quantization (PTQ) of LDMs.

Jialiang Wang, Yan Xia, Ye Yuan

A second-order-based latent factor (SLF) analysis model demonstrates superior performance in graph representation learning, particularly for high-dimensional and incomplete (HDI) interaction data, by incorporating the curvature information of the loss landscape. However, its objective function is commonly bi-linear and non-convex, causing the SLF model to suffer from a low convergence rate. To address this issue, this paper proposes a PID controller-incorporated SLF (PSLF) model, leveraging two key strategies: a) refining learning error estimation by incorporating the PID controller principles, and b) acquiring second-order information insights through Hessian-vector products. Experimental results on multiple HDI datasets indicate that the proposed PSLF model outperforms four state-of-the-art latent factor models based on advanced optimizers regarding convergence rates and generalization performance.

Philippe Hansen-Estruch, David Yan, Ching-Yao Chung et al.

Visual tokenization via auto-encoding empowers state-of-the-art image and video generative models by compressing pixels into a latent space. Although scaling Transformer-based generators has been central to recent advances, the tokenizer component itself is rarely scaled, leaving open questions about how auto-encoder design choices influence both its objective of reconstruction and downstream generative performance. Our work aims to conduct an exploration of scaling in auto-encoders to fill in this blank. To facilitate this exploration, we replace the typical convolutional backbone with an enhanced Vision Transformer architecture for Tokenization (ViTok). We train ViTok on large-scale image and video datasets far exceeding ImageNet-1K, removing data constraints on tokenizer scaling. We first study how scaling the auto-encoder bottleneck affects both reconstruction and generation -- and find that while it is highly correlated with reconstruction, its relationship with generation is more complex. We next explored the effect of separately scaling the auto-encoders' encoder and decoder on reconstruction and generation performance. Crucially, we find that scaling the encoder yields minimal gains for either reconstruction or generation, while scaling the decoder boosts reconstruction but the benefits for generation are mixed. Building on our exploration, we design ViTok as a lightweight auto-encoder that achieves competitive performance with state-of-the-art auto-encoders on ImageNet-1K and COCO reconstruction tasks (256p and 512p) while outperforming existing auto-encoders on 16-frame 128p video reconstruction for UCF-101, all with 2-5x fewer FLOPs. When integrated with Diffusion Transformers, ViTok demonstrates competitive performance on image generation for ImageNet-1K and sets new state-of-the-art benchmarks for class-conditional video generation on UCF-101.

Jialiang Wang, Junzhou Wang, Xin Liao

High-dimensional and incomplete (HDI) data, characterized by massive node interactions, have become ubiquitous across various real-world applications. Second-order latent factor models have shown promising performance in modeling this type of data. Nevertheless, due to the bilinear and non-convex nature of the SLF model's objective function, incorporating a damping term into the Hessian approximation and carefully tuning associated parameters become essential. To overcome these challenges, we propose a new approach in this study, named the adaptive cubic regularized second-order latent factor analysis (ACRSLF) model. The proposed ACRSLF adopts the two-fold ideas: 1) self-tuning cubic regularization that dynamically mitigates non-convex optimization instabilities; 2) multi-Hessian-vector product evaluation during conjugate gradient iterations for precise second-order information assimilation. Comprehensive experiments on two industrial HDI datasets demonstrate that the ACRSLF converges faster and achieves higher representation accuracy than the advancing optimizer-based LFA models.

Hao Wu, Jialiang Wang

Quality-of-Service (QoS) data plays a crucial role in cloud service selection. Since users cannot access all services, QoS can be represented by a high-dimensional and incomplete (HDI) matrix. Latent factor analysis (LFA) models have been proven effective as low-rank representation techniques for addressing this issue. However, most LFA models rely on first-order optimizers and use L2-norm regularization, which can lead to lower QoS prediction accuracy. To address this issue, this paper proposes a double regularized second-order latent factor (DRSLF) model with two key ideas: a) integrating L1-norm and L2-norm regularization terms to enhance the low-rank representation performance; b) incorporating second-order information by calculating the Hessian-vector product in each conjugate gradient step. Experimental results on two real-world response-time QoS datasets demonstrate that DRSLF has a higher low-rank representation capability than two baselines.

Yuewei Yang, Jialiang Wang, Xiaoliang Dai et al.

Diffusion Models (DMs) utilize an iterative denoising process to transform random noise into synthetic data. Initally proposed with a UNet structure, DMs excel at producing images that are virtually indistinguishable with or without conditioned text prompts. Later transformer-only structure is composed with DMs to achieve better performance. Though Latent Diffusion Models (LDMs) reduce the computational requirement by denoising in a latent space, it is extremely expensive to inference images for any operating devices due to the shear volume of parameters and feature sizes. Post Training Quantization (PTQ) offers an immediate remedy for a smaller storage size and more memory-efficient computation during inferencing. Prior works address PTQ of DMs on UNet structures have addressed the challenges in calibrating parameters for both activations and weights via moderate optimization. In this work, we pioneer an efficient PTQ on transformer-only structure without any optimization. By analysing challenges in quantizing activations and weights for diffusion transformers, we propose a single-step sampling calibration on activations and adapt group-wise quantization on weights for low-bit quantization. We demonstrate the efficiency and effectiveness of proposed methods with preliminary experiments on conditional image generation.

Cho-Ying Wu, Jialiang Wang, Michael Hall et al.

The majority of prior monocular depth estimation methods without groundtruth depth guidance focus on driving scenarios. We show that such methods generalize poorly to unseen complex indoor scenes, where objects are cluttered and arbitrarily arranged in the near field. To obtain more robustness, we propose a structure distillation approach to learn knacks from an off-the-shelf relative depth estimator that produces structured but metric-agnostic depth. By combining structure distillation with a branch that learns metrics from left-right consistency, we attain structured and metric depth for generic indoor scenes and make inferences in real-time. To facilitate learning and evaluation, we collect SimSIN, a dataset from simulation with thousands of environments, and UniSIN, a dataset that contains about 500 real scan sequences of generic indoor environments. We experiment in both sim-to-real and real-to-real settings, and show improvements, as well as in downstream applications using our depth maps. This work provides a full study, covering methods, data, and applications aspects.

Bichen Wu, Chaojian Li, Hang Zhang et al.

Neural Architecture Search (NAS) has been widely adopted to design accurate and efficient image classification models. However, applying NAS to a new computer vision task still requires a huge amount of effort. This is because 1) previous NAS research has been over-prioritized on image classification while largely ignoring other tasks; 2) many NAS works focus on optimizing task-specific components that cannot be favorably transferred to other tasks; and 3) existing NAS methods are typically designed to be "proxyless" and require significant effort to be integrated with each new task's training pipelines. To tackle these challenges, we propose FBNetV5, a NAS framework that can search for neural architectures for a variety of vision tasks with much reduced computational cost and human effort. Specifically, we design 1) a search space that is simple yet inclusive and transferable; 2) a multitask search process that is disentangled with target tasks' training pipeline; and 3) an algorithm to simultaneously search for architectures for multiple tasks with a computational cost agnostic to the number of tasks. We evaluate the proposed FBNetV5 targeting three fundamental vision tasks -- image classification, object detection, and semantic segmentation. Models searched by FBNetV5 in a single run of search have outperformed the previous stateof-the-art in all the three tasks: image classification (e.g., +1.3% ImageNet top-1 accuracy under the same FLOPs as compared to FBNetV3), semantic segmentation (e.g., +1.8% higher ADE20K val. mIoU than SegFormer with 3.6x fewer FLOPs), and object detection (e.g., +1.1% COCO val. mAP with 1.2x fewer FLOPs as compared to YOLOX).

Jialiang Wang, Todd Zickler

Localizing stereo boundaries and predicting nearby disparities are difficult because stereo boundaries induce occluded regions where matching cues are absent. Most modern computer vision algorithms treat occlusions secondarily (e.g., via left-right consistency checks after matching) or rely on high-level cues to improve nearby disparities (e.g., via deep networks and large training sets). They ignore the geometry of stereo occlusions, which dictates that the spatial extent of occlusion must equal the amplitude of the disparity jump that causes it. This paper introduces an energy and level-set optimizer that improves boundaries by encoding occlusion geometry. Our model applies to two-layer, figure-ground scenes, and it can be implemented cooperatively using messages that pass predominantly between parents and children in an undecimated hierarchy of multi-scale image patches. In a small collection of figure-ground scenes curated from Middlebury and Falling Things stereo datasets, our model provides more accurate boundaries than previous occlusion-handling stereo techniques. This suggests new directions for creating cooperative stereo systems that incorporate occlusion cues in a human-like manner.

Jialiang Wang, Varun Jampani, Deqing Sun et al.

End-to-end deep learning methods have advanced stereo vision in recent years and obtained excellent results when the training and test data are similar. However, large datasets of diverse real-world scenes with dense ground truth are difficult to obtain and currently not publicly available to the research community. As a result, many algorithms rely on small real-world datasets of similar scenes or synthetic datasets, but end-to-end algorithms trained on such datasets often generalize poorly to different images that arise in real-world applications. As a step towards addressing this problem, we propose to incorporate prior knowledge of scene geometry into an end-to-end stereo network to help networks generalize better. For a given network, we explicitly add a gradient-domain smoothness prior and occlusion reasoning into the network training, while the architecture remains unchanged during inference. Experimentally, we show consistent improvements if we train on synthetic datasets and test on the Middlebury (real images) dataset. Noticeably, we improve PSM-Net accuracy on Middlebury from 5.37 MAE to 3.21 MAE without sacrificing speed.

Jialiang Wang, Todd Zickler

Localizing stereo boundaries is difficult because matching cues are absent in the occluded regions that are adjacent to them. We introduce an energy and level-set optimizer that improves boundaries by encoding the essential geometry of occlusions: The spatial extent of an occlusion must equal the amplitude of the disparity jump that causes it. In a collection of figure-ground scenes from Middlebury and Falling Things stereo datasets, the model provides more accurate boundaries than previous occlusion-handling techniques.

Zhun Deng, Cynthia Dwork, Jialiang Wang et al.

We provide a general framework for characterizing the trade-off between accuracy and robustness in supervised learning. We propose a method and define quantities to characterize the trade-off between accuracy and robustness for a given architecture, and provide theoretical insight into the trade-off. Specifically we introduce a simple trade-off curve, define and study an influence function that captures the sensitivity, under adversarial attack, of the optima of a given loss function. We further show how adversarial training regularizes the parameters in an over-parameterized linear model, recovering the LASSO and ridge regression as special cases, which also allows us to theoretically analyze the behavior of the trade-off curve. In experiments, we demonstrate the corresponding trade-off curves of neural networks and how they vary with respect to factors such as number of layers, neurons, and across different network structures. Such information provides a useful guideline to architecture selection.