From c0bc31e55b20a795cdd6bafd6b6e05cd38f1bcf2 Mon Sep 17 00:00:00 2001 From: kz Date: Fri, 16 Sep 2022 20:47:59 -0700 Subject: [PATCH] update news --- README.md | 19 ++++++++++++++++++- 1 file changed, 18 insertions(+), 1 deletion(-) diff --git a/README.md b/README.md index d3557fc..882fbc5 100644 --- a/README.md +++ b/README.md @@ -6,6 +6,12 @@ This repo include the implementaions of AMSolver, VLMbench, and 6D-CLIPort. +## News + +### 09/16/2022 + +- The work has been accepted by NeurIPS 2022 (Datasets and Benchmarks) ! + ## AMSolver Install Users can use AMSolver to run the current tasks in the VLMbench or build new tasks. In order to run the AMSolver, you should install [Coppliasim 4.1.0](https://www.coppeliarobotics.com/previousVersions) and [PyRep](https://github.com/stepjam/PyRep) first. Then, lets install AMSolver: @@ -74,4 +80,15 @@ python tools/test_config_generator.py ## Add new objects and tasks All object models are saved in *vlm/object_models*. To import new objects into vlmbench, users can use "vlm/object_models/save_model.py". We recommand users first save the object models as a coppeliasim model file (.ttm), then use the extra_from_ttm function inside the save_model.py. More examples can be found in save_model.py. -All tasks templates in the current vlmbench can be found in *vlm/tasks*. To generate new task templates, users can use "tools/task_builder_NLP".py for basic task template generation. Then, the varations of the task can be written as the child classes of the basic task template. More details can refer the codes of *vlm/tasks*. \ No newline at end of file +All tasks templates in the current vlmbench can be found in *vlm/tasks*. To generate new task templates, users can use "tools/task_builder_NLP".py for basic task template generation. Then, the varations of the task can be written as the child classes of the basic task template. More details can refer the codes of *vlm/tasks*. + +## Citation + +``` +@inproceedings{zheng2022vlmbench, + author = {Zheng, Kaizhi and Chen, Xiaotong and Jenkins, Odest Chadwicke and Wang, Xin Eric}, + booktitle = {Proceedings of the Neural Information Processing Systems Track on Datasets and Benchmarks}, + title = {VLMbench: A Compositional Benchmark for Vision-and-Language Manipulation}, + year = {2022} +} +``` \ No newline at end of file