-
Notifications
You must be signed in to change notification settings - Fork 5
Open
Description
Hey everybody,
it seems as if there is an artifact generated by GPT4 during the multi-modal annotation generation in the dataset description of the subset ufba-425 (README.dataset.txt).
Seems like GPT didn't follow the instructions, because classes are missing and it just copied instructions like [distinctive visual features].
# Overview
- [Introduction](#introduction)
- [Object Classes](#object-classes)
- [11](#class-11)
- [12](#class-12)
- [13](#class-13)
- [...]
- [48](#class-48)
# Introduction
The dataset is designed to address a specific object detection task featuring 32 distinct classes. Each class represents a unique type of object that can be identified and annotated in the provided images. The goal is to annotate these objects accurately with bounding boxes, taking into consideration their unique visual characteristics and the instructions for each class.
# Object Classes
## Class 11
### Description
The objects in class 11 are characterized by [distinctive visual features]. They have [describe shape, texture, or any notable characteristics].
Reactions are currently unavailable
Metadata
Metadata
Assignees
Labels
No labels