supplier of weighing and packaging equipment for trade, industries, and logistics.

A science student who majored in computer science at Visvesvaraya University in India. In her graduation research, I developed a website that identifies animal images by image recognition using CNN, which is one of deep learning. The research uses python, Project Jupiter, and TensorFlow. After graduating, she engaged in Android mobile development as a web development intern.



Visvesvaraya Technological University
Computer Science


My grandfather used to have Japanese friends to whom he used to send handwritten letters from India and he used to study Japanese as a hobby. My relatives who worked with a Japanese firm always used to give me positive feedback and encouraged me to go forward if given opportunity. I always wanted to work in abroad and learn a new language. When I got both of these opportunities in Japan during my 4th year of engineering, I was super excited and went for it. I am grateful for the language teachers who helped me with Japanese which helped me clear the exams. After the Japanese exam results I was eagerly waiting for the interviews so that I could go to japan and work. But due to coronavirus the borders were closed and there were no opportunities in japan. I had enrolled with a particular agency which had promised a job but they did not give any clear responses. During this time I came across a former student from Fast offer international on my Instagram feed who gave good response and details about the process for the interviews through fast offer . I registered with fast offer and the first round was an introductory meeting in which they discussed about our interest in working in japan and current working conditions. I did not clear the first round and I was disheartened. Then I started to apply to jobs in japan through LinkedIn but nothing was working for me. After few months I thought I will try again one last time or else I would have to discontinue with my dream of working in japan and start looking for jobs in india.
Finally this time I cleared the first round and I was applicable for the interviews. Soon after I sent my self introduction videos and documents I got shortlisted for one of the companies. I had few mentoring sessions and orientation sessions which really gave me the confidence to attend the interviews but sadly I was not able to clear this interview. I did not give up instead I started practising my Japanese , kept myself busy by doing internships. In the month of November 2022 I was shortlisted for this product based company in Japan. I thought that this is the final chance and I need to give my best. After this I was allotted with a mentor, who helped me and guided me throughout . I will be forever grateful for him. He made sure that all my doubts were cleared and checked each and every presentations. After the mentoring sessions I had orientation session with one of the ASIA to JAPAN team members who again helped me with my presentations and gave extra tips for interviews. I had my first interview with the HR of the company and I cleared the first interview. After the HR round there was a coding round and I passed the interview and accepted the offer. This was one the best year end gift that I could ever asked for and I will remember forever. I am thankful for the entire FAST OFFER International team for giving me this opportunity, I would like to thank the entire team for helping and guiding me. I cannot wait to start my journey in japan.




3) 目的: – 様々な課題と機会の特定(調査)。
– 様々なクラスの動物のデータ集める。
– 動物の種類を識別し、分類する効率的なモデルの開発。
– 動物を識別するためのユーザインタフェースを提供。

4) プロジェクト内容: 動物認識モデルは、主にリアルタイムの野生動物監視に使用することができます。
 視覚的特徴は、動物の分類と識別に重要な役割を果たします。ディープラーニングは、マシンビジョンシステムの研究領域で急速に拡大しているアプリケーションです。ロバストな分類・識別モデルの開発に役立ちます。これは、動物種の分類と識別に使用できるモデルを構築するための利点として使用することができます。多くのデータが生成されるため、VGG16の標準的なアーキテクチャを使用したCNN(Convolution Neural Network)を使用し、正確なCNNを内蔵しています。

5) 使用したソフトウエア: pythonとTensorflowとJupyter Notebook

6) 過程: このプロジェクトで使用するデータセットは、Kaggleから収集したものです。取得したデータセットは、10クラスに属する28,266枚の画像から構成されています。このデータセットには、象、馬、バイソン、猫、犬などの一般的な動物が含まれています。データセット内の画像は全てRGB色空間であり、JPEG形式で保存されています。

7) 結果: 約10エポックの学習で最大83.18%の精度が得られ、テストでは最大76.3%、検証では75.77%でありました。また、入力された動物画像を分類するプロセスをナビゲートするために、Tkinterを用いたユーザフレンドリーなGUIを開発した。また、Tkinter を用いて、入力された動物画像を分類するためのユーザフレンドリな GUI を開発しました。


Read More

Computer Science
Savitribai Phule Pune University
Electronics and Telecommunication
Savitribai Phule Pune University
Electronics and Telecommunication
Universiti Teknologi Malaysia
Bandung Institute of Technology