REAL VAQT TASVIR TASNIFI VA OPTIMALLASHTIRISH TEXNIKALARI
Article Sidebar
Main Article Content
Аннотация:
Ushbu tezis real vaqt tasvir tasnifi texnologiyalari, chuqur o‘rganishga asoslangan modellar va ularni amaliy tizimlarda samarali ishlatish uchun qo‘llaniladigan optimallashtirish texnikalarini tahlil qiladi. Mobil qurilmalar, video monitoring tizimlari, robototexnika va sanoat avtomatizatsiyasi uchun tasvirlarni yuqori tezlikda va aniqlik bilan tasniflash talab etiladi. Shu sababli, konvolyutsion neyron tarmoqlar, yengil arxitekturalar, model siqish, kvantizatsiya, pruning, distillatsiya, ONNX Runtime, TensorRT kabi texnologiyalarning roli batafsil yoritiladi.
Article Details
Как цитировать:
Библиографические ссылки:
Howard, A. et al. (2017). MobileNets: Efficient Convolutional Neural Networks for Mobile Vision Applications.
Tan, M., & Le, Q. (2019). EfficientNet: Rethinking Model Scaling for Convolutional Neural Networks.
Han, S. et al. (2016). Deep Compression: Compressing Deep Neural Networks with Pruning, Trained Quantization and Huffman Coding.
NVIDIA Corporation. (2022). TensorRT Developer Guide.
Google Research Team. (2020). TensorFlow Lite Model Optimization Toolkit.
Microsoft. (2021). ONNX Runtime: High-Performance Machine Learning Inference.