Expert Data Annotator for Complex Engineering CT Scans (US Native)
We are seeking an elite, detail-oriented Data Annotator to join our team for a critical project involving the segmentation and classification of 2D battery slice images.
This is not a simple bounding box task. You will be working with complex, rule-based guidelines to produce highly accurate and consistent annotations that are crucial for training our machine learning models. We are looking for a meticulous professional who thrives on precision and can master a complex set of instructions.
Key Responsibilities:
You will be responsible for two primary annotation tasks on sets of CT scan images:
1. Core Segmentation & Buckling Classification:
- Perform precise polygon segmentation of the battery "core" by identifying the interior boundary of the "first cathode wrap."
- Apply specific geometric rules, such as identifying start/end points based on locally minimal distances, to define the segmentation boundary.
- Accurately handle and annotate complex edge cases, including cathode tip bending, delamination, boundary discontinuities, and tapering, following detailed visual guides.
- Ensure all segmentation masks are smooth, clean, and accurately follow the specified contours.
- Classify each image for "cathode buckling" by identifying the presence of negative curvature on the cathode coil, while correctly ignoring delamination and anode buckling.
2. Overhang Segmentation:
- Create distinct segmentation masks for "cathode" and "anode overhang" components.
- Follow specific inclusion/exclusion rules, labeling only the overhangs that are adjacent to or between cathode segments.
- Maintain a high degree of overlap and accuracy between related masks.
- Leverage the sequential nature of the dataset by efficiently labeling the first image in a series, then copying and fine-tuning subsequent images for maximum productivity.
What We're Looking For (Required Skills):
- Proven Expertise in Complex Segmentation: You must have extensive experience with intricate polygon/mask annotation. Please provide examples or detailed descriptions of past projects.
- Meticulous Attention to Detail: The difference between a correct and incorrect annotation in this project is subtle. You must have an exceptional eye for detail and a commitment to precision.
- Ability to Master Technical Guidelines: You must be able to read, fully comprehend, and meticulously apply a complex, rule-based instruction document with numerous edge cases.
- Excellent Visual Acuity: Experience working with grayscale, low-contrast, or scientific/medical imagery is essential.
- Efficiency and Reliability: You are self-motivated, can manage your time effectively to meet deadlines, and understand workflows designed to improve efficiency (like the copy/paste/refine method described).
- Native English Proficiency (US): Clear communication is key. This role is open to US-based native English speakers only.
Highly Desirable (Brownie Points):
- Experience with medical imaging (CT, MRI) or other forms of engineering/scientific imagery.
- Familiarity with annotation platforms like V7 Darwin, Labelbox, or similar advanced tools.
- A background or interest in engineering, physics, or a related technical field.
Apply Now
Apply Now
This is not a simple bounding box task. You will be working with complex, rule-based guidelines to produce highly accurate and consistent annotations that are crucial for training our machine learning models. We are looking for a meticulous professional who thrives on precision and can master a complex set of instructions.
Key Responsibilities:
You will be responsible for two primary annotation tasks on sets of CT scan images:
1. Core Segmentation & Buckling Classification:
- Perform precise polygon segmentation of the battery "core" by identifying the interior boundary of the "first cathode wrap."
- Apply specific geometric rules, such as identifying start/end points based on locally minimal distances, to define the segmentation boundary.
- Accurately handle and annotate complex edge cases, including cathode tip bending, delamination, boundary discontinuities, and tapering, following detailed visual guides.
- Ensure all segmentation masks are smooth, clean, and accurately follow the specified contours.
- Classify each image for "cathode buckling" by identifying the presence of negative curvature on the cathode coil, while correctly ignoring delamination and anode buckling.
2. Overhang Segmentation:
- Create distinct segmentation masks for "cathode" and "anode overhang" components.
- Follow specific inclusion/exclusion rules, labeling only the overhangs that are adjacent to or between cathode segments.
- Maintain a high degree of overlap and accuracy between related masks.
- Leverage the sequential nature of the dataset by efficiently labeling the first image in a series, then copying and fine-tuning subsequent images for maximum productivity.
What We're Looking For (Required Skills):
- Proven Expertise in Complex Segmentation: You must have extensive experience with intricate polygon/mask annotation. Please provide examples or detailed descriptions of past projects.
- Meticulous Attention to Detail: The difference between a correct and incorrect annotation in this project is subtle. You must have an exceptional eye for detail and a commitment to precision.
- Ability to Master Technical Guidelines: You must be able to read, fully comprehend, and meticulously apply a complex, rule-based instruction document with numerous edge cases.
- Excellent Visual Acuity: Experience working with grayscale, low-contrast, or scientific/medical imagery is essential.
- Efficiency and Reliability: You are self-motivated, can manage your time effectively to meet deadlines, and understand workflows designed to improve efficiency (like the copy/paste/refine method described).
- Native English Proficiency (US): Clear communication is key. This role is open to US-based native English speakers only.
Highly Desirable (Brownie Points):
- Experience with medical imaging (CT, MRI) or other forms of engineering/scientific imagery.
- Familiarity with annotation platforms like V7 Darwin, Labelbox, or similar advanced tools.
- A background or interest in engineering, physics, or a related technical field.
Apply Now
Apply Now