Project AURIONAI Use-Cases Research & Innovation
UC Berkeley
vision

DLA (Deep Layer Aggregation)

Backbone that aggregates features across layers, from the paper "Deep Layer Aggregation"; DLA-34 is the CenterNet backbone.

UC BerkeleyUC BerkeleyUSA
Figure from the DLA (Deep Layer Aggregation) paper: Different approaches to aggregation.
Levelmodel
Classvision
Parameters26.4M
Contextnot in the notes
Released20 Jul 2017
Opennessopen weights permissive
Runs onmobile or NPU, CPU server
Statusactive

R&D Topics

Image classification backbone architectures (CNN and vision transformer)produced by

Identity

Developer
UC Berkeley
Family
not in the notes
Released
20 Jul 2017 sourcesannouncement page
Weights published
24 Apr 2023
Record checked
12 Oct 2026
Where this comes from
Developer and dates from the research notes. Release date is the vendor's public announcement; weights date is when open weights were published.

Architecture and Size

Architecture
CNN
Total parameters
26.4M sources
Active parameters
–
Checkpoint parameters
–
Context length
not in the notes
Input
image
Output
not in the notes
Where this comes from
Parameter counts as the developer states them. Where the checkpoint holds a different total, both are shown (register A16).

Openness and Access

Openness
open weights permissive
Licence
BSD 3-Clause sources
Access routes
open weights download
Where this comes from
The licence file governs over registry tags and aggregator labels.

Deployment

Runs on
mobile or NPU, CPU server
Footprint
Edge or toolkit builds listed in: Qualcomm AI Hub, OpenVINO.
Runtimes
not in the notes
Where this comes from
Deployment tiers and runtimes as named in the notes. Footprint is the notes' figure, not a measurement by AURION.

Indonesia

Indonesian support
not stated
Regional languages
not in the notes
Jakarta hosting
unknown
Languages
not in the notes
Where this comes from
Indonesia fields are carried on every model; unknown is an allowed value.

Benchmark Results

The notes record no benchmark result for this model.

Lineage

The notes record no lineage for this model.

Industry Use Cases

No industry use case links this model.

Sources

  1. arxiv.org/abs/1707.06484 https://arxiv.org/abs/1707.06484
  2. huggingface.co/timm/dla102x.in1k https://huggingface.co/timm/dla102x.in1k
  3. aihub.qualcomm.com/models/dla102x https://aihub.qualcomm.com/models/dla102x
  4. github.com/openvinotoolkit/open_model_zoo/tree/master/mo... https://github.com/openvinotoolkit/open_model_zoo/tree/master/models/public/dla-34
  5. github.com/huggingface/pytorch-image-models/blob/main/RE... https://github.com/huggingface/pytorch-image-models/blob/main/README.md
  6. github.com/onnx/models/tree/main/Computer_Vision/dla102_... https://github.com/onnx/models/tree/main/Computer_Vision/dla102_Opset16_timm
  7. arxiv.org/html/1707.06484 https://arxiv.org/html/1707.06484

To cite this page: Project AURION, AI Use-Cases Research & Innovation, "DLA (Deep Layer Aggregation)", https://aurion.semai.id/models/dla/, records of October 2026.