Search Results for "grid"

...In this work, we present a new framework termed BEVFormer, which learns unified BEV representations with spatiotemporal transformers to support multiple autonomous driving perception tasks. In a nutshell, BEVFormer exploits both spatial and temporal information by interacting with spatial and temporal space through predefined grid-shaped BEV queries. To aggregate spatial information, we design spatial cross-attention that each BEV query extracts the spatial features from the regions of interest across camera views. For temporal information, we propose temporal self-attention to recurrently fuse the history BEV information. Our approach achieves the new state-of-the-art 56.9\% in terms of NDS metric on the nuScenes \texttt{test} set, which is 9.0 points higher than previous best arts and on par with the performance of LiDAR-based baseline.

Torch Points 3D is a framework for developing and testing common deep learning models to solve tasks related to unstructured 3D spatial data i.e. Point Clouds. The framework currently integrates some of the best-published architectures and it integrates the most common public datasets for ease of reproducibility. It heavily relies on Pytorch Geometric and Facebook Hydra library thanks for the great work! We aim to build a tool that can be used for benchmarking SOTA models, while also...

Bowtie is a library for writing dashboards in Python. No need to know web frameworks or JavaScript, focus on building functionality in Python. Interactively explore your data in new ways! Deploy and share with others! Bowtie uses Yarn to manage node packages. If you installed Bowtie through conda, Yarn was also installed as a dependency. Yarn can be installed through conda. An early integration with Jupyter has been prototyped. I encourage you to try it out and share feedback. I hope this...

...The dashboard layout and blueprint styles are pre-packaged, and provide only the essentials, while getting out of the way. Drag-and-drop your layout easily and intuitively. Multiple layout modes - bootstrap grid-based or totally freeform. Data utilities for munging and manipulating data suitable for various charts.

You assemble *blocks* in your django template, with each block wired to a tastypie resource, which in turn is mapped to a django model. Each *block* will reside in a dijit ContentPane which you must define in your template. You have a choice of 2 layouts for each *block*, Form or Grid based. Form displays a single record in either a 1, 2 or 3 column layout while Grid displays many rows in a spreadsheet like grid. You can perform edits in place on fields in both layouts. You can define the master-detail relationship between the *blocks* and enblok will automatically display the related detail *block* rows as the user navigates the master *block*'s rows. ...

Console-based tool that will allow for multiple types of automation testing to run in parallel. Written in Python