Skip to main content

​
Sequence-Scoped Methods

​
client.annotation_issues.list_by_sequence()

List all annotation issues for a specific sequence.
​
sequence_uid
str
required
Unique identifier of the sequence
​
dataset_item_uid
str
Filter by dataset item UID
​
project_uid
str
Filter by project UID
​
response
List[AnnotationIssue]
List of annotation issues for the sequence
from avala import Avala

client = Avala(api_key="your-api-key")

issues = client.annotation_issues.list_by_sequence(
    sequence_uid="seq-123",
    project_uid="proj-456"
)
for issue in issues:
    print(f"{issue.uid}: {issue.description}")

​
client.annotation_issues.create()

Create a new annotation issue for a sequence.
​
sequence_uid
str
required
Unique identifier of the sequence
​
tool_uid
str
required
UID of the annotation tool associated with the issue
​
problem_uid
str
required
UID identifying the type of problem
​
dataset_item_uid
str
Dataset item where the issue occurs
​
project_uid
str
Project context for the issue
​
priority
str
Issue priority level
​
severity
str
Issue severity level
​
description
str
Detailed description of the issue
​
wrong_class
str
Incorrect class label
​
correct_class
str
Correct class label
​
should_re_annotate
bool
Whether the annotation should be redone
​
should_delete
bool
Whether the annotation should be deleted
​
frames_affected
str
Description of which frames are affected
​
coordinates
Any
Spatial coordinates related to the issue
​
query_params
Dict[str, Any]
Additional query parameters
​
object_uid
str
UID of the object related to the issue
​
response
AnnotationIssue
The newly created annotation issue
issue = client.annotation_issues.create(
    sequence_uid="seq-123",
    tool_uid="tool-456",
    problem_uid="problem-789",
    priority="high",
    severity="critical",
    description="Incorrect bounding box placement",
    wrong_class="car",
    correct_class="truck",
    should_re_annotate=True
)

​
client.annotation_issues.update()

Update an existing annotation issue.
​
sequence_uid
str
required
Unique identifier of the sequence
​
issue_uid
str
required
Unique identifier of the annotation issue
​
status
str
Updated status of the issue
​
priority
str
Updated priority level
​
severity
str
Updated severity level
​
description
str
Updated description
​
tool_uid
str
Updated tool UID
​
problem_uid
str
Updated problem UID
​
wrong_class
str
Updated wrong class label
​
frames_affected
str
Updated frames affected description
​
response
AnnotationIssue
The updated annotation issue
updated_issue = client.annotation_issues.update(
    sequence_uid="seq-123",
    issue_uid="issue-789",
    status="resolved",
    priority="low"
)

​
client.annotation_issues.delete()

Delete an annotation issue.
​
sequence_uid
str
required
Unique identifier of the sequence
​
issue_uid
str
required
Unique identifier of the annotation issue to delete
client.annotation_issues.delete(
    sequence_uid="seq-123",
    issue_uid="issue-789"
)

​
Dataset-Scoped Methods

​
client.annotation_issues.list_by_dataset()

List all annotation issues for a dataset.
​
owner
str
required
Username or organization slug
​
dataset_slug
str
required
Dataset identifier slug
​
sequence_uid
str
Filter by sequence UID
​
response
List[AnnotationIssue]
List of annotation issues for the dataset
issues = client.annotation_issues.list_by_dataset(
    owner="username",
    dataset_slug="my-dataset",
    sequence_uid="seq-123"
)

​
client.annotation_issues.get_metrics()

Retrieve annotation issue metrics for a dataset.
​
owner
str
required
Username or organization slug
​
dataset_slug
str
required
Dataset identifier slug
​
sequence_uid
str
Filter metrics by sequence UID
​
response
AnnotationIssueMetrics
Metrics summary for annotation issues
metrics = client.annotation_issues.get_metrics(
    owner="username",
    dataset_slug="my-dataset"
)
print(f"Total issues: {metrics.total_count}")

​
Tool Methods

​
client.annotation_issues.list_tools()

List available annotation quality control tools for a dataset type.
​
dataset_type
str
required
Type of dataset (e.g., “image”, “video”, “lidar”)
​
response
List[AnnotationIssueToolDetail]
List of available QC tools with their problem types
tools = client.annotation_issues.list_tools(
    dataset_type="image"
)
for tool in tools:
    print(f"{tool.name}: {tool.description}")
    for problem in tool.problems:
        print(f"  - {problem.name}")

​
Async Methods

All methods are available in async form via client.annotation_issues when using AsyncAvala: 
from avala import AsyncAvala

async with AsyncAvala(api_key="your-api-key") as client:
    # Sequence-scoped
    issues = await client.annotation_issues.list_by_sequence(
        sequence_uid="seq-123"
    )
    
    issue = await client.annotation_issues.create(
        sequence_uid="seq-123",
        tool_uid="tool-456",
        problem_uid="problem-789",
        description="Issue description"
    )
    
    updated = await client.annotation_issues.update(
        sequence_uid="seq-123",
        issue_uid="issue-789",
        status="resolved"
    )
    
    await client.annotation_issues.delete(
        sequence_uid="seq-123",
        issue_uid="issue-789"
    )
    
    # Dataset-scoped
    dataset_issues = await client.annotation_issues.list_by_dataset(
        owner="username",
        dataset_slug="my-dataset"
    )
    
    metrics = await client.annotation_issues.get_metrics(
        owner="username",
        dataset_slug="my-dataset"
    )
    
    # Tools
    tools = await client.annotation_issues.list_tools(
        dataset_type="image"
    )

Build docs developers (and LLMs) love

Get started for freeTalk to us