More Like this

1. A Snapshot-based Approach for Self-supervised Feature Learning and Weakly-supervised Classification on Point Cloud Data

   https://doi.org/10.5220/0010230103990408  

   Li, Xingye; Zhu, Zhigang (January 2021, VISAPP 2021, the 16th International Conference on Computer Vision Theory and Applications)

   null (Ed.)

   Manually annotating complex scene point cloud datasets is both costly and error-prone. To reduce the reliance on labeled data, we propose a snapshot-based self-supervised method to enable direct feature learning on the unlabeled point cloud of a complex 3D scene. A snapshot is defined as a collection of points sampled from the point cloud scene. It could be a real view of a local 3D scan directly captured from the real scene, or a virtual view of such from a large 3D point cloud dataset. First the snapshots go through a self-supervised pipeline including both part contrasting and snapshot clustering for feature learning. Then a weakly-supervised approach is implemented by training a standard SVM classifier on the learned features with a small fraction of labeled data. We evaluate the weakly-supervised approach for point cloud classification by using varying numbers of labeled data and study the minimal numbers of labeled data for a successful classification. Experiments are conducted on three public point cloud datasets, and the results have shown that our method is capable of learning effective features from the complex scene data without any labels. 

   more » « less
2. Self-supervised Feature Learning by Cross-modality and Cross-view Correspondences

   Jing, L; Zhang, L; Tian, Y (June 2021, The 4th Multimodal Learning and Applications (MULA) Workshop in conjunction with CVPR 2021)

   null (Ed.)

   The success of supervised learning requires large-scale ground truth labels which are very expensive, time- consuming, or may need special skills to annotate. To address this issue, many self- or un-supervised methods are developed. Unlike most existing self-supervised methods to learn only 2D image features or only 3D point cloud features, this paper presents a novel and effective self-supervised learning approach to jointly learn both 2D image features and 3D point cloud features by exploiting cross-modality and cross-view correspondences without using any human annotated labels. Specifically, 2D image features of rendered images from different views are extracted by a 2D convolutional neural network, and 3D point cloud features are extracted by a graph convolution neural network. Two types of features are fed into a two-layer fully connected neural network to estimate the cross-modality correspondence. The three networks are jointly trained (i.e. cross-modality) by verifying whether two sampled data of different modalities belong to the same object, meanwhile, the 2D convolutional neural network is additionally optimized through minimizing intra-object distance while maximizing inter-object distance of rendered images in different views (i.e. cross-view). The effectiveness of the learned 2D and 3D features is evaluated by transferring them on five different tasks including multi-view 2D shape recognition, 3D shape recognition, multi-view 2D shape retrieval, 3D shape retrieval, and 3D part-segmentation. Extensive evaluations on all the five different tasks across different datasets demonstrate strong generalization and effectiveness of the learned 2D and 3D features by the proposed self-supervised method. 

   more » « less
3. “Canopy fingerprints” for characterizing three-dimensional point cloud data of soybean canopies

   https://doi.org/10.3389/fpls.2023.1141153  

   Young, Therin J.; Jubery, Talukder Z.; Carley, Clayton N.; Carroll, Matthew; Sarkar, Soumik; Singh, Asheesh K.; Singh, Arti; Ganapathysubramanian, Baskar (March 2023, Frontiers in Plant Science)

   Advances in imaging hardware allow high throughput capture of the detailed three-dimensional (3D) structure of plant canopies. The point cloud data is typically post-processed to extract coarse-scale geometric features (like volume, surface area, height, etc.) for downstream analysis. We extend feature extraction from 3D point cloud data to various additional features, which we denote as ‘canopy fingerprints’. This is motivated by the successful application of the fingerprint concept for molecular fingerprints in chemistry applications and acoustic fingerprints in sound engineering applications. We developed an end-to-end pipeline to generate canopy fingerprints of a three-dimensional point cloud of soybean [Glycine max(L.) Merr.] canopies grown in hill plots captured by a terrestrial laser scanner (TLS). The pipeline includes noise removal, registration, and plot extraction, followed by the canopy fingerprint generation. The canopy fingerprints are generated by splitting the data into multiple sub-canopy scale components and extracting sub-canopy scale geometric features. The generated canopy fingerprints are interpretable and can assist in identifying patterns in a database of canopies, querying similar canopies, or identifying canopies with a certain shape. The framework can be extended to other modalities (for instance, hyperspectral point clouds) and tuned to find the most informative fingerprint representation for downstream tasks. These canopy fingerprints can aid in the utilization of canopy traits at previously unutilized scales, and therefore have applications in plant breeding and resilient crop production. 

   more » « less
4. Act3D: 3D Feature Field Transformers for Multi-Task Robotic Manipulation

   Theophile Gervet, Zhou Xiao (November 2023, Conference on Robot Learning)

   3D perceptual representations are well suited for robot manipulation as they easily encode occlusions and simplify spatial reasoning. Many manipulation tasks require high spatial precision in end-effector pose prediction, which typically demands high-resolution 3D feature grids that are computationally expensive to process. As a result, most manipulation policies operate directly in 2D, foregoing 3D inductive biases. In this paper, we introduce Act3D, a manipulation policy transformer that represents the robot’s workspace using a 3D feature field with adaptive resolutions dependent on the task at hand. The model lifts 2D pre-trained features to 3D using sensed depth, and attends to them to compute features for sampled 3D points. It samples 3D point grids in a coarse to fine manner, featurizes them using relative-position attention, and selects where to focus the next round of point sampling. In this way, it efficiently computes 3D action maps of high spatial resolution. Act3D sets a new state-of-the-art in RLBench, an established manipulation benchmark, where it achieves 10% absolute improvement over the previous SOTA 2D multi-view policy on 74 RLBench tasks and 22% absolute improvement with 3x less compute over the previous SOTA 3D policy. We quantify the importance of relative spatial attention, large-scale vision-language pre-trained 2D backbones, and weight tying across coarse-to-fine attentions in ablative experiments. Code and videos are available at our project site: https://act3d.github.io/. 

   more » « less
5. Diverse task-driven modeling of macaque V4 reveals functional specialization towards semantic tasks

   https://doi.org/10.1371/journal.pcbi.1012056  

   Cadena, Santiago A; Willeke, Konstantin F; Restivo, Kelli; Denfield, George; Sinz, Fabian H; Bethge, Matthias; Tolias, Andreas S; Ecker, Alexander S (May 2024, PLOS Computational Biology)

   Einhäuser, Wolfgang (Ed.)

   Responses to natural stimuli in area V4—a mid-level area of the visual ventral stream—are well predicted by features from convolutional neural networks (CNNs) trained on image classification. This result has been taken as evidence for the functional role of V4 in object classification. However, we currently do not know if and to what extent V4 plays a role in solving other computational objectives. Here, we investigated normative accounts of V4 (and V1 for comparison) by predicting macaque single-neuron responses to natural images from the representations extracted by 23 CNNs trained on different computer vision tasks including semantic, geometric, 2D, and 3D types of tasks. We found that V4 was best predicted by semantic classification features and exhibited high task selectivity, while the choice of task was less consequential to V1 performance. Consistent with traditional characterizations of V4 function that show its high-dimensional tuning to various 2D and 3D stimulus directions, we found that diverse non-semantic tasks explained aspects of V4 function that are not captured by individual semantic tasks. Nevertheless, jointly considering the features of a pair of semantic classification tasks was sufficient to yield one of our top V4 models, solidifying V4’s main functional role in semantic processing and suggesting that V4’s selectivity to 2D or 3D stimulus properties found by electrophysiologists can result from semantic functional goals. 

   more » « less