Object Detection Using Yolov11 Model
Introduction This project builds on Ultralytics repository, utilizing YOLOv11 model for object detection on COCO dataset, which contains 122,000 natural images across 80 object classes. Using Tensorleap, we analyze the model’s latent space, uncover hidden patterns, identify mislabeled samples, in order to improve model performance and dataset quality. Tensorleap Platform Tensorleap is a platform for […]
Action Recognition – Kinetics-400
Introduction Video classification is a common and complex area in machine learning. Unlike other vision tasks that contain only spatial information (Segmentation, Object Detection, Mono-Depth), the addition of the temporal dimension makes error analysis and pinpointing failures root-cause especially challenging. In this blog, we explore that challenge through the lens of Tensorleap, a powerful platform […]
The MNIST Dataset
In this example, we used the MNIST dataset (handwritten digits) to train a convolutional neural network (CNN) for image classification, and reached a validation accuracy of ~0.98 on our trained model.
Enhancing ADAS Object Detection with YOLOv7: Overcoming Data Shifts through Latent Space Analysis and Efficient Data Synthesis
Tensorleap enables efficient exploration of a model’s latent space to detect data shifts in unseen samples, including unlabeled data. This repository showcases how to detect and address such issues, specifically within an object detection task for ADAS.
LibriSpeech Dataset With Wav2vec
A speech recognition task for predicting the text from an audio file. We utilize a wav2vec model trained on the librispeech-asr benchmark, implemented using Keras and Tensorflow.
GLPN: Monocular Depth Estimation
Depth estimation task using a single image to predict depth.
Sentiment Analysis
The IMDB dataset is one of the most popular sentiment analysis datasets. It contains multiple movie reviews, each annotated with either a positive or negative label. In this example, a classifier model was built to predict positive and negative reviews.
CelebA Dataset Gender and Age Group Prediction with MobileNetV2
This project showcases a deep learning classification task for predicting the gender and age group of face individuals in images. We train a MobileNetV2 model to perform this task on CelebA benchmark using Keras and Tensorflow implementation.
Detecting Pneumonia – Chest X-Ray Images
The project uses transfer learning from a pre-trained DensNet201 network to classify chest X-ray images into one of three classes- normal, bacteria, or virus. In a DenseNet, we use dense connections between layers, using Dense Blocks, where each layer (with identical feature maps) directly connects with each other through dense connections. On the left panel of the chest X-ray […]
Optical Flow – Raft
This project utilizes the RAFT (Recurrent All-pairs Field Transforms) model that estimates Optical Flow using the KITTI dataset.