Real-time perception, or streaming perception, is a crucial aspect of autonomous driving that has yet to be thoroughly explored in existing research. To address this gap, we present DAMO-StreamNet, an optimized framework that combines recent advances from the YOLO series with a comprehensive analysis of spatial and temporal perception mechanisms, delivering a cutting-edge solution. The key innovations of DAMO-StreamNet are: (1) A robust neck structure incorporating deformable convolution, enhancing the receptive field and feature alignment capabilities. (2) A dual-branch structure that integrates short-path semantic features and long-path temporal features, improving motion state prediction accuracy. (3) Logits-level distillation for efficient optimization, aligning the logits of teacher and student networks in semantic space. (4) A real-time forecasting mechanism that updates support frame features with the current frame, ensuring seamless streaming perception during inference.~Our experiments demonstrate that DAMO-StreamNet surpasses existing state-of-the-art methods, achieving 37.8% (normal size (600, 960)) and 43.3% (large size (1200, 1920)) sAP without using extra data. This work not only sets a new benchmark for real-time perception but also provides valuable insights for future research. Additionally, DAMO-StreamNet can be applied to various autonomous systems, such as drones and robots, paving the way for real-time perception.
Model | size | velocity | sAP 0.5:0.95 |
sAP50 | sAP75 | coco pretrained models | weights |
---|---|---|---|---|---|---|---|
DAMO-StreamNet-S | 600×960 | 1x | 31.8 | 52.3 | 31.0 | link | link |
DAMO-StreamNet-M | 600×960 | 1x | 35.5 | 57.0 | 36.2 | link | link |
DAMO-StreamNet-L | 600×960 | 1x | 37.8 | 59.1 | 38.6 | link | link |
DAMO-StreamNet-L | 1200×1920 | 1x | 43.3 | 66.1 | 44.6 | link | link |
Please find the teacher model here.
You can refer to StreamYOLO/LongShortNet to install the whole environments.
Refer to here to prepare the Argoverse-HD dataset. Please put the dataset into ./data
or create a symbolic links to the dataset in ./data
.
Please download the models provided above into ./models
and organize them as:
./models
├── checkpoints
│ ├── streamnet_l_1200x1920.pth
│ ├── streamnet_l.pth
│ ├── streamnet_m.pth
│ └── streamnet_s.pth
├── coco_pretrained_models
│ ├── yolox_l_drfpn.pth
│ ├── yolox_m_drfpn.pth
│ └── yolox_s_drfpn.pth
└── teacher_models
└── l_s50_still_dfp_flip_ep8_4_gpus_bs_8
└── best_ckpt.pth
bash run_train.sh
bash run_eval.sh
Our implementation is mainly based on StreamYOLO and LongShortNet. We gratefully thank the authors for their wonderful works.