In this article, we will explore how to use the Gemini CLI to develop a 2D English word game by integrating it with Context 7 and Task Master AI. This setup allows for efficient task management and access to the latest documentation, enhancing the development process.
Find the game demo here: Word Cloud Shooter
Continue ReadingWhen developing software—especially as codebases grow to thousands or even tens of thousands of lines—developers often encounter hard-to-fix bugs, or fixing one issue may introduce new problems. Even simple projects can require significant time to implement all desired features. Many existing AI programming tools, such as Cursor, are limited by their context window, typically supporting only about 10K tokens, which is roughly 400 to 600 lines of standard Python code. This short context length restricts the AI’s understanding of the entire project, making it difficult to identify key dependencies and architectural information. As a result, these tools can only handle single files or small-scale projects.
Continue ReadingIn this post, I’ll share how I built a classic 2048 game from scratch by combining the Vibe Coding workflow with Roo Code’s AI-powered collaborative development features. Through real project steps, practical tips, and personal notes, you’ll get a glimpse of what modern AI-assisted development feels like — and how it points toward a future where developers and AI co-create software seamlessly.
Check out the live demo here: https://game-2048.hoohoo.top/
Vibe Coding is an emerging development methodology that emphasizes visualization, real-time interaction, and AI collaboration. The goal is to lower the barrier to entry for coding, shorten development cycles, and improve code quality all at once.
Continue Reading最近在研究語音辨識，順帶看了一下 Python，這裡大概就介紹一下在學習 Python 深度學習推薦入門的幾個 Libraries。
Python 運用在深度學習算相當具有優勢，支援的 libraries 相當豐富，在這裡推薦幾個 Libraries ，進行介紹。
在了解及初次熟悉這些 Libraries 時，建議可以透過以下方式：
如果是透過 Anaconda 啟用 Jupyther Notebook ，則可以參考這部影片 https://www.youtube.com/watch?v=-MyjG00la2k
Continue Reading最近在研究語音辨識，順帶看了一下 Python，這裡大概就介紹一下在學習 Python 資料科學計算推薦入門的幾個 Libraries。
Python 運用在科學資料計算相當具有優勢，支援的 libraries 相當豐富，在這裡推薦幾個 Libraries ，進行介紹。
在了解及初次熟悉這些 Libraries 時，建議可以透過以下方式：
如果是透過 Anaconda 啟用 Jupyther Notebook ，則可以參考這部影片 https://www.youtube.com/watch?v=-MyjG00la2k
Continue Reading在執行 Python Keras過程，如果出現錯誤提示：
ModuleNotFoundError: No module named ‘keras.models’; ‘keras’ is not a package
如果中間過程提示缺少套件，缺什麼就安裝什麼
pip3 install keras
可以查看自己目前安裝的套件有哪些
pip3 list
如果安裝後仍出現 No module named 'eras.models'… 那這時請確認一下是否你有將檔案名稱命名為 keras.py
如果有，這表示 Python 在執行過程會認為你是要將這個檔案 import
因此，解決方式是將檔案名稱更換掉就可以
Continue Reading最近在嘗試幾種音波比對的方向，大致測試過幾種技術，
這裡主要紀錄音波轉成 MFCC 之後，進行實測的一些過程及觀察數據，
首先，根據不同狀況錄製了不同的音頻，接著就直接針對音頻進行比較，
（免責聲明！以下測試並無使用正規統計方式評測，僅供參考）
Continue Reading本篇主要針對 AudioDiff 在音檔比對的實測
Continue Reading下載AudioCompare
git clone https://github.com/charlesconnell/AudioCompare.git
進行比對
./audiomatch -f howareudoing.wav -f howareudo.wav
# MATCH howareudoing.wav howareudo.wav (39)
./audiomatch -f howareudoing.wav -f howsgoingtoday.wav
# MATCH howareudoing.wav howsgoingtoday.wav (13)
./audiomatch -f howareudoing.wav -f howareudoing.wav
# MATCH howareudoing.wav howareudoing.wav (112)
./audiomatch -f application.wav -f howareudo.wav
# NO MATCH
目前測試結果，準確性相當低
透過文字自動生成的音檔，可比對出匹配指數 但是透過實際發音的檔案，同樣唸法仍無法匹配出來
Continue Reading安裝 EPEL 來使用其他 repo 資源
yum install -y epel-release
sudo rpm –import /etc/pki/rpm-gpg/RPM-GPG-KEY-EPEL-7
sudo rpm –import http://li.nux.ro/download/nux/RPM-GPG-KEY-nux.ro
sudo rpm -Uvh http://li.nux.ro/download/nux/dextop/el7/x86_64/nux-dextop-release-0-1.el7.nux.noarch.rpm
yum repolist
接著執行
yum install -y ffmpeg
查看版本
ffmpeg -version
解決 Python [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed
最近在測試 wit.ai 語音辨識過程出現 CERTIFICATE_VERIFY_FAILED 錯誤，原因是 Python 3.6 之後的版本更改了 OpenSSL 協定，因此需要手動進行更新
Could not request results ; recognition connection failed: [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate (_ssl.c:1056) Could not request results from Wit.ai service; recognition connection failed: [SSL: CERTIFICATE_VERIFY_FAILED] certificate verify failed: unable to get local issuer certificate (_ssl.c:1056)
Continue Reading執行
pip install PyAudio
如果出現以下次錯誤
Collecting PyAudio
Downloading https://files.pythonhosted.org/packages/ab/42/b4f04721c5c5bfc196ce156b3c768998ef8c0ae3654ed29ea5020c749a6b/PyAudio-0.2.11.tar.gz
Building wheels for collected packages: PyAudio
Building wheel for PyAudio (setup.py) ... error
ERROR: Complete output from command /opt/rh/rh-python36/root/usr/bin/python3 -u -c 'import setuptools, tokenize;__file__='"'"'/tmp/pip-install-aqdgoicv/PyAudio/setup.py'"'"';f=getattr(tokenize, '"'"'open'"'"', open)(__file__);code=f.read().replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, __file__, '"'"'exec'"'"'))' bdist_wheel -d /tmp/pip-wheel-3dqdycqf --python-tag cp36:
ERROR: running bdist_wheel
running build
running build_py
原因在於，需要安裝 portaudio ，解決方式如下:
wget http://portaudio.com/archives/pa_stable_v190600_20161030.tgz
tar -xvf pa_stable_v190600_20161030.tgz
cd portaudio
./configure
接著在執行安裝，應該就會正常
pip install PyAudio
這裡找了一下 AWS 跟 Google 語音辨識的計費比較．
目前評估兩家換算的金額都很接近，
如果使用一次單字平台測驗，其中包括單字加上老師補充單字
以每次使用 30次及 60 次口說題，大致會花費 $NT 6~11 塊
| AWS | ||
|---|---|---|
| 產品名稱 | Transcribe | speech-to-text |
| 網址 | https://aws.amazon.com/tw/transcribe/ | https://cloud.google.com/speech-to-text/pricing?hl=zh-tw |
| 計費方式 | 0~15秒 0.18，之後每秒 0.012 | 每 15秒 0.06 |
| 測驗一個單字 | 0.18 | 0.18 |
| 單字平台（30次) | 5.4 | 5.4 |
| 單字平台（60次) | 10.8 | 10.8 |
| 單位 台幣 |
備註
中國區域語音辨識價格約略在 0.004 元/次 人民幣 驰声 http://www.chivox.com/englishoralevaluation/index.aspx 百度 http://ai.baidu.com/docs#/ASR-Pricing/top 讯飞 https://www.xfyun.cn/services/voicedictation
Continue Reading