skip to main content

Explore Research Products in the NSF-PAR

Title: Towards Long-Tailed 3D Detection
Contemporary autonomous vehicle (AV) benchmarks have advanced techniques for training 3D detectors, particularly on large-scale lidar data. Surprisingly, although semantic class labels naturally follow a long-tailed distribution, contemporary benchmarks focus on only a few common classes (e.g., pedestrian and car) and neglect many rare classes in-the-tail (e.g., debris and stroller). However, AVs must still detect rare classes to ensure safe operation. Moreover, semantic classes are often organized within a hierarchy, e.g., tail classes such as child and construction-worker are arguably subclasses of pedestrian. However, such hierarchical relationships are often ignored, which may lead to misleading estimates of performance and missed opportunities for algorithmic innovation. We address these challenges by formally studying the problem of Long-Tailed 3D Detection (LT3D), which evaluates on all classes, including those in-the-tail. We evaluate and innovate upon popular 3D detection codebases, such as CenterPoint and PointPillars, adapting them for LT3D. We develop hierarchical losses that promote feature sharing across common-vs-rare classes, as well as improved detection metrics that award partial credit to "reasonable" mistakes respecting the hierarchy (e.g., mistaking a child for an adult). Finally, we point out that fine-grained tail class accuracy is particularly improved via multimodal fusion of RGB images with LiDAR; simply put, small fine-grained classes are challenging to identify from sparse (lidar) geometry alone, suggesting that multimodal cues are crucial to long-tailed 3D detection. Our modifications improve accuracy by 5% AP on average for all classes, and dramatically improve AP for rare classes (e.g., stroller AP improves from 3.6 to 31.6)! Our code is available at this https URL.  more » « less
Award ID(s):
1950811
NSF-PAR ID:
10441693
Author(s) / Creator(s):
; ; ;
Date Published:
Journal Name:
International Conference on Robot Learning
Format(s):
Medium: X
Sponsoring Org:
National Science Foundation
More Like this
  1. ; ; ; ( , IEEE Conference on Computer Vision and Pattern Recognition)
    null (Ed.)
    Vehicle detection with visual sensors like lidar and camera is one of the critical functions enabling autonomous driving. While they generate fine-grained point clouds or high-resolution images with rich information in good weather conditions, they fail in adverse weather (e.g., fog) where opaque particles distort lights and significantly reduce visibility. Thus, existing methods relying on lidar or camera experience significant performance degradation in rare but critical adverse weather conditions. To remedy this, we resort to exploiting complementary radar, which is less impacted by adverse weather and becomes prevalent on vehicles. In this paper, we present Multimodal Vehicle Detection Network (MVDNet), a two-stage deep fusion detector, which first generates proposals from two sensors and then fuses region-wise features between multimodal sensor streams to improve final detection results. To evaluate MVDNet, we create a procedurally generated training dataset based on the collected raw lidar and radar signals from the open-source Oxford Radar Robotcar. We show that the proposed MVDNet surpasses other state-of-the-art methods, notably in terms of Average Precision (AP), especially in adverse weather conditions. The code and data are available at https://github.com/qiank10/MVDNet. 
    more » « less
  2. ; ; ; ; ; ; ; ( , Advances in neural information processing systems)
    Vanilla models for object detection and instance segmentation suffer from the heavy bias toward detecting frequent objects in the long-tailed setting. Existing methods address this issue mostly during training, e.g., by re-sampling or re- weighting. In this paper, we investigate a largely overlooked approach — post- processing calibration of confidence scores. We propose NORCAL, Normalized Calibration for long-tailed object detection and instance segmentation, a simple and straightforward recipe that reweighs the predicted scores of each class by its training sample size. We show that separately handling the background class and normalizing the scores over classes for each proposal are keys to achieving superior performance. On the LVIS dataset, NORCAL can effectively improve nearly all the baseline models not only on rare classes but also on common and frequent classes. Finally, we conduct extensive analysis and ablation studies to offer insights into various modeling choices and mechanisms of our approach. Our code is publicly available at https://github.com/tydpan/NorCal. 
    more » « less
  3. ; ; ; ( , 2021 IEEE Winter Conference on Applications of Computer Vision (WACV))
    For the task of image classification, researchers work arduously to develop the next state-of-the-art (SOTA) model, each bench-marking their own performance against that of their predecessors and of their peers. Unfortunately, the metric used most frequently to describe a model’s performance, average categorization accuracy, is often used in isolation. As the number of classes increases, such as in fine-grained visual categorization (FGVC), the amount of information conveyed by average accuracy alone dwindles. While its most glaring weakness is its failure to describe the model’s performance on a class-by-class basis, average accuracy also fails to describe how performance may vary from one trained model of the same architecture, on the same dataset, to another (both averaged across all categories and at the per-class level). We first demonstrate the magnitude of these variations across models and across class distributions based on attributes of the data, comparing results on different visual domains and different per-class image distributions, including long-tailed distributions and few-shot subsets. We then analyze the impact various FGVC methods have on overall and per-class variance. From this analysis, we both highlight the importance of reporting and comparing methods based on information beyond overall accuracy, as well as point out techniques that mitigate variance in FGVC results. 
    more » « less
  4. ; ; ; ; ; ( , European Conference on Computer Vision)
    One fundamental challenge in building an instance segmen- tation model for a large number of classes in complex scenes is the lack of training examples, especially for rare objects. In this paper, we ex- plore the possibility to increase the training examples without laborious data collection and annotation. We find that an abundance of instance segments can potentially be obtained freely from object-centric images, according to two insights: (i) an object-centric image usually contains one salient object in a simple background; (ii) objects from the same class often share similar appearances or similar contrasts to the background. Motivated by these insights, we propose a simple and scalable frame- work FreeSeg for extracting and leveraging these “free” object fore- ground segments to facilitate model training in long-tailed instance seg- mentation. Concretely, we investigate the similarity among object-centric images of the same class to propose candidate segments of foreground instances, followed by a novel ranking of segment quality. The resulting high-quality object segments can then be used to augment the exist- ing long-tailed datasets, e.g., by copying and pasting the segments onto the original training images. Extensive experiments show that FreeSeg yields substantial improvements on top of strong baselines and achieves state-of-the-art accuracy for segmenting rare object categories. Our code is publicly available at https://github.com/czhang0528/FreeSeg. 
    more » « less
  5. ; ; ; ; ; ; ; ; ( , Plant Methods)
    Abstract Background 3D imaging, such as X-ray CT and MRI, has been widely deployed to study plant root structures. Many computational tools exist to extract coarse-grained features from 3D root images, such as total volume, root number and total root length. However, methods that can accurately and efficiently compute fine-grained root traits, such as root number and geometry at each hierarchy level, are still lacking. These traits would allow biologists to gain deeper insights into the root system architecture. Results We present TopoRoot, a high-throughput computational method that computes fine-grained architectural traits from 3D images of maize root crowns or root systems. These traits include the number, length, thickness, angle, tortuosity, and number of children for the roots at each level of the hierarchy. TopoRoot combines state-of-the-art algorithms in computer graphics, such as topological simplification and geometric skeletonization, with customized heuristics for robustly obtaining the branching structure and hierarchical information. TopoRoot is validated on both CT scans of excavated field-grown root crowns and simulated images of root systems, and in both cases, it was shown to improve the accuracy of traits over existing methods. TopoRoot runs within a few minutes on a desktop workstation for images at the resolution range of 400^3, with minimal need for human intervention in the form of setting three intensity thresholds per image. Conclusions TopoRoot improves the state-of-the-art methods in obtaining more accurate and comprehensive fine-grained traits of maize roots from 3D imaging. The automation and efficiency make TopoRoot suitable for batch processing on large numbers of root images. Our method is thus useful for phenomic studies aimed at finding the genetic basis behind root system architecture and the subsequent development of more productive crops. 
    more » « less
  • Citation Formats
  • MLA
  • APA
  • Chicago
  • BibTeX
  • Save / Share this Record
  • Send to Email