Hello, I encountered a model download issue while following the emotion example in SpacemiT’s official model-zoo-vision repository, and I’d like to ask for help.
Repository URL:
Relevant directories:
examples/emotion
applications/emotion_detection
According to the configuration file, the emotion example requires the model at:
~/.cache/models/vision/emotion/emotion_resnet50_final.q.onnx
In addition, applications/emotion_detection uses:
~/.cache/models/vision/emotion/emotion_resnet50_features.dynq.onnx
~/.cache/models/vision/emotion/emotion_lstm_affwild2_final.q.onnx
However, after running the official download script, I received XML error files of just over 400 bytes instead of ONNX models. Here’s what I found:
emotion_resnet50_final.q.onnx: 443 bytes, XML 1.0 document
emotion_resnet50_features.dynq.onnx: 449 bytes, XML 1.0 document
emotion_lstm_affwild2_final.q.onnx: 448 bytes, XML 1.0 document
The files contain:
NoSuchKey
The specified key does not exist.
The corresponding download URLs are:
https://archive.spacemit.com/spacemit-ai/model_zoo/vision/emotion/emotion_resnet50_final.q.onnx
https://archive.spacemit.com/spacemit-ai/model_zoo/vision/emotion/emotion_lstm_affwild2_final.q.onnx
Have these model files been moved to a new download location? Could you provide working download URLs or recommend emotion model files for the current version?
I’d also like to confirm a few configuration details:
- Is the class order in assets/labels/emotion.txt still:
neutral, happiness, sadness, surprise, fear, disgust, anger - Is the configuration in examples/emotion/config/emotion.yaml still compatible with the current model?
- Does enabling SpaceMITExecutionProvider on the K1 require any additional environment variables or shared library path configuration?
- If libonnxruntime.so cannot be found at runtime, what is the recommended way to set LD_LIBRARY_PATH?
Thank you.
:~/model-zoo-vision$ cd ~/model-zoo-vision
rm -rf ~/.cache/models/vision/emotion_test
mkdir -p ~/.cache/models/vision/emotion_test
cd ~/.cache/models/vision/emotion_test
echo “下载 static emotion model…”
curl -L -o emotion_resnet50_final.q.onnx
https://archive.spacemit.com/spacemit-ai/model_zoo/vision/emotion/emotion_resnet50_final.q.onnx
echo “下载 feature model…”
doneho -c 120 “$f”====“节:”/spacemit-ai/model_zoo/vision/emotion/emotion_ls
Downloading static emotion model…
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:–:-- --:–:-- --:–:-- 0 0 0 0 0 0 0 0 --:–:-- --:–:-- --:–:-- 100 443 100 443 0 0 778 0 --:–:-- --:–:-- --:–:-- 778
Downloading feature model…
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:–:-- --:–:-- --:–:-- 100 449 100 449 0 0 886 0 --:–:-- --:–:-- --:–:-- 8100 449 100 449 0 0 885 0 --:–:-- --:–:-- --:–:-- 885
Downloading LSTM model…
% Total % Received % Xferd Average Speed Time Time Time Current
Dload Upload Total Spent Left Speed
0 0 0 0 0 0 0 0 --:–:-- --:–:-- --:–:-- 100 448 100 448 0 0 813 0 --:–:-- --:–:-- --:–:-- 816
Check results:
total 12K
-rw-rw-r-- 1 bianbu1 bianbu1 448 Jun 11 17:03 emotion_lstm_affwild2_final.q.onnx
-rw-rw-r-- 1 bianbu1 bianbu1 449 Jun 11 17:03 emotion_resnet50_features.dynq.onnx
-rw-rw-r-- 1 bianbu1 bianbu1 443 Jun 11 17:03 emotion_resnet50_final.q.onnx
emotion_lstm_affwild2_final.q.onnx: XML 1.0 document, ASCII text
emotion_resnet50_features.dynq.onnx: XML 1.0 document, ASCII text
emotion_resnet50_final.q.onnx: XML 1.0 document, ASCII text
First 120 bytes of each file:
===== emotion_lstm_affwild2_final.q.onnx =====
NoSuchKey
The specified key does not exist.
NoSuchKey
The specified key does not exist.
NoSuchKey
The specified key does not exist.</Me
bianbu1@k1:~/.cache/models/vision/emotion_test$ rm -rf ~/.cache/models/vision/emotion_test
rm -f ~/.cache/models/vision/emotion/*.onnx
bianbu1@k1:~/.cache/models/vision/emotion_test$ cd /home/bianbu1
pkill -f optimized_camera
pkill -f decision_engine.py
pkill -f web_preview.py
pkill -f screen_interaction.py
pkill -f emotion_ai_worker.py
pkill -f aplay
/home/bianbu1/start_warm_station_all.sh
sleep 10
cat /home/bianbu1/camera_status.json
ps -o pid,pcpu,pmem,cmd -C python3us | head -c 500
Stopping old programs…
Starting camera recognition program…
Starting decision management program…
Enhanced state analysis program is temporarily disabled
Starting small-screen interaction program…
Starting web preview program…
Warm Heart Station multimodal recognition system started
Open in a browser on your computer:
http://10.34.218.232:8080
View camera log: tail -f /home/bianbu1/camera.log
View enhanced analysis log: tail -f /home/bianbu1/enhanced.log
View web log: tail -f /home/bianbu1/web.log
{
“user_present”: true,
“emotion”: “NEUTRAL”,
“confidence”: 0.93,
“probabilities”: {
“HAPPY”: 0.07,
“NEUTRAL”: 0.93,
“STANDBY”: 0.0
},
“face_box”: {
“x”: 140,
“y”: 14,
“w”: 150,
“h”: 150
},
“time”: “2026-06-11 17:05:10”
}{“user_present”: true, “emotion”: “NEUTRAL”, “confidence”: 0.93, “probabilities”: {“HAPPY”: 0.07, “NEUTRAL”: 0.93, “STANDBY”: 0.0}, “face_box”: {“x”: 140, “y”: 14, “w”: 150, “h”: 150}, “time”: “2026-06-11 17:05:11”, “last_action”: “”, “chat_reply”: “The current state is relatively calm. You can ask the user if they need help.”, “chat_state”: “Rule-based emotion Q&A connected”, “raw_emotion”: “NEUTRAL”, “display_probabilities”: {“HAPPY”: 0.07, “NEUTRAL”: 0.93, “STANDBY”: 0.0}, “current_frame_probabilities”: {"HA
PID %CPU %MEM CMD
32068 40.3 1.7 python3 optimized_camera_working.py
32070 1.6 0.1 python3 -u decision_engine.py
32072 69.5 3.6 python3 -u screen_interaction.py
32073 4.7 0.2 python3 web_preview.py
I want to integrate the modei-zoo model into my board.