Few-Shot Object Detection Dataset (FSOD) is a high-diverse dataset specifically designed for few-shot object detection and intrinsically designed to evaluate thegenerality of a model on novel categories.

Overview

• Introduction
• Object Classes
  • Bobcat
  • Cattle
  • Ocelot
  • Opossum

Introduction

This dataset is designed to solve the task of detecting and classifying various wildlife and domestic animals from images. The classes included are Bobcat, Cattle, Ocelot, and Opossum. Each class will be introduced with a brief description of its distinctive features.

Object Classes

Bobcat

Description

Bobcats are medium-sized wildcats characterized by a stocky build, tufted ears, and distinctive facial ruff. They often have a spotted coat and a short "bobbed" tail.

Instructions

• Annotate the entire visible body of the bobcat, including limbs and tail.
• Ensure the bounding box encompasses all extremities without excluding any part of the visible body.
• Do not include obscured parts if less than 20% of the animal is visible.

Cattle

Description

Cattle are large domestic animals with robust bodies and short legs. They have a generally smooth coat and distinctive broad snouts.

Instructions

• Capture the full body within the bounding box, ensuring all visible parts of the head, body, and legs are included.
• Pay attention to the ears and snouts, as these are distinctive features.
• Avoid annotating if the animal is not sufficiently visible to clearly identify as cattle.

Ocelot

Description

Ocelots are small wildcats known for their sleek bodies and striking rosette-patterned fur. They have rounded ears and a long tail.

Instructions

• Draw the bounding box around the full visible extent of the ocelot, focusing on capturing its rosette-patterned fur.
• Make sure to include the tail and ears within the annotation.
• Parts that are heavily obscured should be excluded unless enough is visible for clear identification.

Opossum

Description

Opossums are small to medium-sized marsupials with pointed snouts, stiff whiskers, and typically white to grayish fur. Their tails are long and scaly.

Instructions

• Annotate the entire visible body of the opossum, including the head, body, and tail.
• Ensure the bounding box includes the distinctive pointed snout.
• Do not annotate if the visibility is too low to confidently identify the species.

SAR Fsod

## Overview

SAR Fsod is a dataset for object detection tasks - it contains Ships annotations for 364 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [CC BY 4.0 license](https://creativecommons.org/licenses/CC BY 4.0).

Overview

• Introduction
• Object Classes
  • Enemy
  • Head

Introduction

The dataset is designed to recognize and annotate "enemy" characters and their "heads" in gaming scenarios. The goal is to accurately delineate these classes for object detection tasks.

• Enemy: Represents the full body of an enemy character.
• Head: Represents the head of an enemy character, often a target in gaming scenarios.

Object Classes

Enemy

Description

The "enemy" class refers to the complete visible body of an adversary character. In a gaming context, these are the figures actively engaged as adversaries and often appear in combat poses or scenarios.

Instructions

• Annotate the entire visible body of the enemy character.
• Ensure that the bounding box includes all discernible parts, such as limbs, torso, and head if fully visible.
• Do not include weapons, gear, or other objects that extend beyond the body’s boundaries.
• If an enemy is partially occluded, include only the visible sections that can be confidently identified.

Head

Description

The "head" class targets the head portion of an enemy character. This section often serves as a critical aiming point within first-person shooter games.

Instructions

• Focus on the head, capturing the full extent of its visible structure.
• If the head is turned or partially visible, include only the discernible portion.
• Avoid including parts of the neck or shoulders unless they are indistinguishable from the head structure.
• Ensure the bounding box does not extend into other body parts or non-head-related elements.
• Disambiguate from the enemy class by ensuring the annotation is confined tightly around the head area only.

FSOD Valorant

## Overview

FSOD Valorant is a dataset for object detection tasks - it contains Valorant Agents annotations for 200 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [Public Domain license](https://creativecommons.org/licenses/Public Domain).

Fruitjes Fsod Fhzu

## Overview

Fruitjes Fsod Fhzu is a dataset for object detection tasks - it contains Fruitjes Fsod Fhzu Fhzu annotations for 427 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [MIT license](https://creativecommons.org/licenses/MIT).

All Elements Fsod Sxkf

## Overview

All Elements Fsod Sxkf is a dataset for object detection tasks - it contains All Elements Fsod Sxkf Sxkf annotations for 265 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [MIT license](https://creativecommons.org/licenses/MIT).

Overview

• Introduction
• Object Classes
  • Wheat Heads

Introduction

The dataset focuses on detecting wheat heads in agricultural images to assist with monitoring and analysis. It contains annotations for the following class:

• Wheat Heads: Wheat heads are the parts of the wheat plant where the grains are formed.

Object Classes

Wheat Heads

Description

The wheat head is part of the wheat plant, distinguished by its elongated shape and spikelets containing grains. They may appear amidst leaves and stems, often seen in dense clusters or isolated. Wheat heads may be green or yellow in color depending on their moisture content.

Instructions

• Annotate the entire wheat head, capturing the full elongated shape from the base near the stem to the tip, even if partially obscured by other parts of the plant.
• Ensure the bounding box includes the entire visible head while excluding leaves and stems unless they are integral to the shape.
• Do not annotate if the head is less than 20% visible, or if it is unclear whether the object is a wheat head.
• Avoid annotating overly blurred objects that cannot be confidently identified as wheat heads.
• Cross-referencing other instances can help disambiguate partially visible wheat heads.

Nih Xray Fsod Hhze

## Overview

Nih Xray Fsod Hhze is a dataset for object detection tasks - it contains Nih Xray Fsod Hhze Hhze annotations for 245 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [MIT license](https://creativecommons.org/licenses/MIT).

Overview

• Introduction
• Object Classes
  • Flir Camera Objects
  • Bicycle
  • Car
  • Dog
  • Person

Introduction

This dataset is focused on object detection in images captured by FLIR cameras. The primary objective is to identify and annotate various commonplace objects within images. The dataset consists of five primary classes: - Flir Camera Objects: Indeterminate objects usually characterized by thermal signatures. - Bicycle: Two-wheeled vehicles that are human-powered. - Car: Motorized four-wheeled vehicles. - Dog: Canine animals. - Person: Human figures.

Object Classes

Flir Camera Objects

Description

This category encapsulates objects that appear in thermal images which don't fall neatly into other defined classes. These objects might appear as distinct thermal blobs or patterns not distinctly associated with recognizable objects.

Instructions

Annotate any object that appears as a unique thermal signature which cannot be characterized as a bicycle, car, dog, or person. Exclude areas that blend into the background and cannot be distinctly separated from it.

Bicycle

Description

A bicycle is typically characterized by its two large circular wheels, a frame, and handlebars.

Instructions

Annotate the entire outline of the bicycle, ensuring the bounding box covers both wheels and the frame. Do not include rider unless clearly distinct as a separate object.

Car

Description

Cars appear as four-wheeled motor vehicles, often distinct in shape by a defined outline that includes headlights, windows, and body frame.

Instructions

Encapsulate the entire vehicle within the bounding box, including visible wheels. Exclude any attachments that do not distinctly look like part of the car, such as trailers or external cargo unless consistent with the car's dimensions.

Dog

Description

Dogs are four-legged animals with recognizable features such as a head, tail, and a general body outline.

Instructions

Include the whole outline of the dog, capturing from the head to the tail and all visible limbs. If partially occluded, guess the missing parts within reason.

Person

Description

Persons appear as human silhouettes with visible head, torso, and limbs.

Instructions

Surround the visible silhouette of the person with a bounding box, including the head and visible extremities. Do not box separate clothing items that do not form part of a continuous silhouette.

Ball Qgqhv Fsod Cmkg

## Overview

Ball Qgqhv Fsod Cmkg is a dataset for object detection tasks - it contains Ball Qgqhv Fsod Cmkg Cmkg annotations for 256 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [MIT license](https://creativecommons.org/licenses/MIT).

Overview

• Introduction
• Object Classes
  • Bicycle
  • Car
  • Dog
  • Person

Introduction

This dataset is designed for object detection using images captured by FLIR cameras. It includes four classes: bicycle, car, dog, and person.

• Bicycle: Two-wheeled, pedal-driven vehicles.
• Car: Motor vehicles with four wheels, used for transportation.
• Dog: Domesticated animals, often four-legged and furry.
• Person: Human beings, typically upright in posture.

Object Classes

Bicycle

Description

Bicycles are two-wheeled vehicles that can be identified by their circular wheels and frame structure. They often feature handlebars and pedals.

Instructions

• Annotate the bicycle frame and wheels completely. Ensure both wheels are included, even if partially visible.
• Do not label accessories like water bottles unless they are part of the main structure.
• If a person is riding the bicycle, the person's body should be annotated separately.

Car

Description

Cars are four-wheeled motor vehicles. They are identifiable by their larger structure compared to bicycles, with a distinct hood and trunk area.

Instructions

• Include all visible parts of the car, like the body and wheels.
• Do not separate individual parts of the car unless they are distinct objects, e.g., a detached roof rack.
• Exclude reflections of the car if visible in nearby surfaces.

Dog

Description

Dogs are four-legged animals often seen on roadsides or sidewalks. They are typically identified by their fur and tail.

Instructions

• Annotate the entire body of the dog, ensuring legs, head, and tail are included.
• Do not label dogs if only a small portion is visible, making identification uncertain.
• Avoid annotating any accessories unless they are very evident (e.g., a leash attached to the dog).

Person

Description

People are identified by their upright posture and features like arms and legs. Typically captured engaging in various activities.

Instructions

• Include the whole person, capturing arms, legs, and head.
• Persons in motion, like walking or running, should still be annotated if they are easily distinguishable.
• If a person is partially hidden behind objects, only annotate visible parts that make identification possible.

Flir Camera Objects Fsod Gqhf

## Overview

Flir Camera Objects Fsod Gqhf is a dataset for object detection tasks - it contains Flir Camera Objects Fsod Gqhf Gqhf annotations for 587 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [MIT license](https://creativecommons.org/licenses/MIT).

Paper Parts Fsod Iuph

## Overview

Paper Parts Fsod Iuph is a dataset for object detection tasks - it contains Paper Parts Fsod Iuph Iuph annotations for 924 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [MIT license](https://creativecommons.org/licenses/MIT).

Grass Weeds Fsod Xhtf

## Overview

Grass Weeds Fsod Xhtf is a dataset for object detection tasks - it contains Grass Weeds Fsod Xhtf Xhtf annotations for 264 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [MIT license](https://creativecommons.org/licenses/MIT).

Overview

• Introduction
• Object Classes
  • Author
  • Chapter
  • Equation
  • Equation Number
  • Figure
  • Figure Caption
  • Footnote
  • List of Content Heading
  • List of Content Text
  • Page Number
  • Paragraph
  • Reference Text
  • Section
  • Subsection
  • Subsubsection
  • Table
  • Table Caption
  • Table of Contents Text
  • Title

Introduction

This dataset is designed to annotate the structural elements of academic papers. It aims to train models to recognize different parts of a paper. Each class corresponds to a text or graphical element commonly found in papers.

• Author: The name(s) of the person(s) who wrote the document.
• Chapter: The major divisions within the paper, usually denoted by a number and a title.
• Equation: Mathematical formulas or expressions.
• Equation Number: The numeral identifiers for equations.
• Figure: Visual representations like graphs or charts.
• Figure Caption: Text descriptions associated with figures.
• Footnote: Additional information at the bottom of the page.
• List of Content Heading: The titles of content sections in a list.
• List of Content Text: Descriptions or details within a list of content.
• Page Number: The numeral indicating the page's position.
• Paragraph: Blocks of text conveying an idea or point.
• Reference Text: Citations or bibliographic information.
• Section: Main headings within a chapter.
• Subsection: Subheadings under a section.
• Subsubsection: Further subdivisions under a subsection.
• Table: Data or information arranged in rows and columns.
• Table Caption: Text descriptions associated with tables.
• Table of Contents Text: Entries listing sections and page numbers.
• Title: The main heading or name of the paper.

Object Classes

Author

Description

Text indicating the name(s) of the author(s), typically found near the beginning of a document.

Instructions

Identify the text block containing the author names. It usually follows the title and may include affiliations. Do not include titles, affiliations or titles of sections adjacent to author names.

Chapter

Description

Indicates a major division of the document, often labeled with a number and title.

Instructions

Locate text labeled with "Chapter" followed by a number and title. Capture the entire heading, ensuring no unrelated text is included.

Equation

Description

Symbols and numbers arranged to represent a mathematical concept.

Instructions

Draw boxes around all mathematical expressions, excluding any accompanying text or numbers identifying the equations.

Equation Number

Description

Numerals used to uniquely identify equations.

Instructions

Identify numbers in parentheses next to equations. Do not include equation text or variables.

Figure

Description

Visual content such as graphs, diagrams, code or images.

Instructions

Outline the entire graphical representation. Do not include captions or any surrounding text.

Figure Caption

Description

Text providing a description or explanation above or below a figure.

Instructions

Identify the text directly associated with a figure. Ensure no unrelated figures or text are included.

Footnote

Description

Clarifications or additional details located at the bottom of a page.

Instructions

Locate text at the page's bottom that refers back to a mark or reference in the main text. Exclude any unrelated content.

List of Content Heading

Description

Headings at the list of context text, identifying its purpose or content. This may also be called a list of figures.

Instructions

Identify and label only the heading for lists in content sections. Do not include subsequent list items.

List of Content Text

Description

The detailed entries or points in a list. These often summarize all figures in the paper.

Instructions

Identify each item in a content list. Exclude list headings and any non-list content.

Page Number

Description

Numerical indication of the current page.

Instructions

Locate numbers typically positioned at the top or bottom margins. Do not include text or symbols beside the numbers.

Paragraph

Description

Blocks of text separated by spacing or indentation.

Instructions

Enclose individual text blocks that form coherent sections. Ensure each paragraph is distinguished separately.

Reference Text

Description

Bibliographic information found typically in a reference sect

Overview

• Introduction
• Object Classes
  • Adult
  • Juvenile
  • Piglet

Introduction

This dataset aims to classify different stages of growth within boars. The classes include:

• Adult: Fully grown boars.
• Juvenile: Young boars that are smaller than adults but still sizable.
• Piglet: Very young and small boars.

Object Classes

Adult

Description

Adults are fully grown and large in size, typically taking up a significant portion of the image. They have well-defined features, such as distinct body and facial structures.

Instructions

• Annotate the entire body, ensuring all visible parts are included.
• Do not include shadows or blurry sections that do not distinctly outline the animal.
• Focus only on clear, distinguishable features of adults.

Juvenile

Description

Juveniles are smaller than adults but significantly larger than piglets. They retain the body shape of adults but are not fully grown.

Instructions

• Annotate visible juveniles, focusing on their smaller size relative to adults.
• Ensure the entire juvenile is included, without capturing unrelated nearby objects.
• Distinguish from adults by their notably smaller size and slightly less mature features.

Piglet

Description

Piglets are the smallest, typically found very close to the ground. They have a more compact and less developed body shape compared to adults and juveniles.

Instructions

• Carefully annotate the small, rounded structure of piglets.
• Ensure to capture only the piglet, avoiding any overlap with other classes.
• Distinguish from juveniles by their more compact and infant-like size.

Overview

• Introduction
• Object Classes
  • Attack
  • Block
  • Defense
  • Serve
  • Set
  • Ball

Introduction

This dataset is used for recognizing volleyball actions and objects. It includes six classes: Attack, Block, Defense, Serve, Set, and Ball. The aim is to annotate instances of these actions and objects in volleyball games to assist in automated sports analysis.

Object Classes

Attack

Description

The "Attack" action involves a player attempting to hit the ball forcefully over the net to score a point, usually characterized by an airborne leap and raised arm with the palm facing forward.

Instructions

• Annotate the player as soon as their body is in mid-air with an arm extended above the head with an intention to strike the ball.
• Ensure the bounding box includes the full extent of the player’s body, anticipating motion, but exclude the ball unless it directly touches the player’s hand at the marking moment.

Block

Description

"Block" involves players reaching up near the net with both hands raised above the head to stop or deflect an opponent's attack.

Instructions

• Capture players with both hands above their head, with palms facing the incoming ball.
• Ensure the bounding box contains the player’s arms and hands, particularly positioned near the net area, ready to block.

Defense

Description

"Defense" is characterized by players adopting a low stance, often with forearms parallel to the floor, ready to receive the ball from an opponent’s attack.

Instructions

• Annotate players crouched with arms arranged to form a platform, usually positioned behind the front line.
• Don’t include players in a standing position unrelated to an immediate ball-defense action.

Serve

Description

"Serve" involves a player initiating play by striking the ball from behind the end line to send it over the net.

Instructions

• Identify a player behind the end line, typically poised to toss and hit or having just made contact with the ball.
• Include the full player body preparing to serve, with a focus on hand positioning relative to the ball pre and post-contact.

Set

Description

A "Set" involves a player using their fingertips to push the ball upwards to set up a spike, often positioned right in front of the net.

Instructions

• Box the player with arms elevated and fingers spread, set under the ball, usually just after a teammate's pass.
• Ensure proper focus on hand positions distinctly forming a cup shape under the ball.

Ball

Description

The volleyball is a spherical object used in play, identifiable by distinct panels often rotating mid-flight.

Instructions

• Annotate the ball in every frame where it is visually distinct and not immediately obscured by another player.
• Focus on capturing the ball entirely within the bounding box regardless of its position relative to players or the court.

Stomata Cells Fsod Fmjz

## Overview

Stomata Cells Fsod Fmjz is a dataset for object detection tasks - it contains Stomata Cells Fsod Fmjz Fmjz annotations for 238 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [MIT license](https://creativecommons.org/licenses/MIT).

Pig Detection Kaimq Fsod Arql

## Overview

Pig Detection Kaimq Fsod Arql is a dataset for object detection tasks - it contains Pig Detection Kaimq Fsod Arql Arql annotations for 579 images.

## Getting Started

You can download this dataset for use within your own projects, or fork it into a workspace on Roboflow to create your own model.

  ## License

  This dataset is available under the [MIT license](https://creativecommons.org/licenses/MIT).