v0.2.0-alpha.3

What's Changed

• v0.2.0-alpha1 by @jamjamjon in #185
• update by @jamjamjon in #187
• update by @jamjamjon in #188
• feat: upgrade ort to v2.0.0-rc11 and add NvTensorRT-RTX ep by @jamjamjon in #189
• feat(yolo): add YOLO26 models by @jamjamjon in #193
• fix: CUDA Image Processor device compatibility and performance by @jamjamjon in #205
• Add BiRefNet and YOLOE-26-Prompt-Free models by @jamjamjon in #207
• Add YOLOE-26 prompt-free and prompt-based models by @jamjamjon in #208
• chore: change default static input shapes for coreml by @wep21 in #206
• Change default visual encoder dtype to int4 by @killzoner in #214
• feat: Add YOLOE-26-prompt-based by @jamjamjon in #217
• fix: dynamic batch pose estimation for trt by @wep21 in #216
• Add MoboleGaze model and update docs by @jamjamjon in #223
• Add PP-DocLayout-V1/V2 model by @jamjamjon in #224
• Add RF-DETR models by @jamjamjon in #225
• Update docs by @jamjamjon in #228
• Update trt-ep setting by @jamjamjon in #229
• Add PP-Doclayout-V3 by @jamjamjon in #233
• Add yolo26-det-int8 by @jamjamjon in #235

New Contributors

• @killzoner made their first contribution in #214

Full Changelog: v0.1.11...v0.2.0-alpha.3

v0.1.11 - Sam3-Image & Sam3-tracker

What's Changed

• Add sam3-image model by @jamjamjon in #164
• feat(sam3-image): add TensorRT support by @jamjamjon in #166
• refactor(sam3-image): code refine by @jamjamjon in #169
• Add SAM3 Python export and inference scripts by @jamjamjon in #172
• Add sam3-tracker and refactor style system by @jamjamjon in #177
• Introduce tracing by @jamjamjon in #178
• Split out HTTP(S) (ureq) and image default-formats into features by @glasgow-haskell-omelette in #165

Full Changelog: v0.1.10...v0.1.11

v0.1.10 – GroundingDINO multi-batch inference, mm-gdino/llmdet integration & new class-name post-processing

What's Changed

• feat(gdino): add multi-batch inference support; integrate mm-gdino … by @jamjamjon in #162

Full Changelog: v0.1.9...v0.1.10

New Supported ONNX Model

• gdino-base
• gdino-tiny
• llmdet-base
• llmdet-large
• llmdet-tiny
• mm-gdino-base-all
• mm-gdino-base-o365v1-goldg-v3det
• mm-gdino-large-all
• mm-gdino-large-o365v2-oiv6-goldg
• mm-gdino-tiny-o365v1-goldg-grit-v3det
• mm-gdino-tiny-o365v1-goldg-grit
• mm-gdino-tiny-o365v1-goldg-v3det
• mm-gdino-tiny-o365v1-goldg

v0.1.9 - Add RT-DETRv4

What's Changed

• Make Y elements public and remove fake optimizations by @numfin in #160
• Add RT-DETRv4 by @jamjamjon in #161

New Contributors

• @numfin made their first contribution in #160

Full Changelog: v0.1.8...v0.1.9

Add RAM & RAM++ models

What's Changed

• Add RAM & RAM++ models by @jamjamjon in #158

Full Changelog: v0.1.7...v0.1.8

Optimize dependencies and modularize model features

What's Changed

• chore(deps): bump aksr to 0.0.4 by @jamjamjon in #155
• refactor: optimize dependencies and modularize model features by @jamjamjon in #157

Full Changelog: v0.1.6...v0.1.7

Add MOT support and the RF-DETR-Seg model

What's Changed

• Add ByteTracker by @jamjamjon in #153
• Add RF-DETR-seg model by @jamjamjon in #154

Full Changelog: v0.1.5...v0.1.6

YOLOE now supports both visual-prompt and text-prompt modes.

What's Changed

• Change frames per second (fps) from usize to f32 by @paulingalls in #148
• Add visual prompt support and fix font loading bug by @jamjamjon in #152

Full Changelog: v0.1.4...v0.1.5

v0.0.20

What's Changed

• chore(ort): upgrade ort to v2.0.0-rc.9 by @TD-Sky in #52

New Contributors

• @TD-Sky made their first contribution in #52

Full Changelog: v0.0.19...v0.0.20

v0.0.19

What's Changed

• Add Apple ml-depth-pro model by @jamjamjon in #46

Full Changelog: v0.0.18...v0.0.19

Model	Demo
Depth Anything v1-s (512 x 512)
Depth Anything v2-s (512 x 512)
Depth Pro (1536 x 1536)