Daniel Flower Dot Com Banner


This is a report of a semester long project I did at the University of Auckland between March and July, 2005. You can view the report by viewing the pages in this category or by downloading the PDF below.

Report Abstract

As the number of robots in the world increases, from automatic vacuum cleaners, to toy robot dogs, to autonomous vehicles for the military, the need for effective algorithms to control these agents is becoming increasingly more important. Conventional path finding techniques have relied on having a representation of the world that could be analysed mathematically to find the best path. However, when an agent is placed into the real world in a place it has not seen before, conventional techniques fail and a fundamentally different approach to path finding is required. The agent must rely on its senses, such as the input from a mounted camera, using this information to get around. In this project, a virtual city is implemented, and agents must navigate their way around it by using only what they see from their point of view. By feeding what the agents see into a neural network, they are able to learn how to avoid obstacles, follow the road, and they also show promise in using this technique for path finding.


Neural network, vision based path finding

This is the PDF version of my report.


Comments for this page
neural network path finding
posted by matt on 10/09/2005 10:19:47 p.m. (NZ time)
very interesting indeed. do you plan any follow up work on this topic? like building a robot based on it?
re: neural network path finding
posted by Daniel Flower on 4/10/2005 12:56:12 a.m. (NZ time)
Thanks. Unfortunately I don't have any spare robots at the moment, so maybe later...
posted by Victor on 24/04/2006 5:36:50 p.m. (NZ time)
i've just taken a course on Neural Computation. They currently have a technique to train for infinite layers of neural nets. It is based on Restricted Boltzmann Machine, very interesting.
Actual credit unions with the best savings accounts
posted by best savings accounts on 24/08/2008 1:32:43 a.m. (NZ time)
Coll blog, thanks.
Nice research
posted by eriq on 30/10/2010 12:09:31 p.m. (NZ time)
Very nice and interesting, I have do same research for my postgraduate project using neural network for pathfinding,, but I have plan to use ray casting to percept the input not from camera or vision which you have done. Would you like to share the implementation or source code for my project's reference ? if you will please reply me at eriq[dot]adams[at]gmail[dot]com
thanks alot
posted by Biya on 2/10/2012 3:03:21 a.m. (NZ time)
其實 前途 呢樣野無一個好肯定既定義 原因係好睇天時地利同個人興趣 我而家覺得我自己做緊呢份工好有前途咩 俾你做你可能會覺得一片黑暗 就好似保險咁丫 有 d 人覺得只係頭個幾年辛苦 做多幾年有批客o係手就開始舒服 再俾心機做多幾年就可以得閒同 d 客 high tea 個個月袋十幾皮 好有前途咁喎 但係有 d 人又會可能覺得日日如是同 d 師奶傾湊仔經 大 firm 又唔到你做 細 firm 又得個做字 朋友聽倒你做保險又會避之則吉咁 好厭惡者咁 所以其實揀份有 前途 既工同讀書揀科一樣 切忌邊科 hit 就讀邊科 或者人做乜你做乜 為錢揀人工高既咩 世界變得太快 幾年之後個行飽和唔通你又轉 而且為錢而做既工 如果你唔鐘意 就正如施永青先生所講 工作不快樂 生活不快樂 唔開心既工 你一定無心機做 自然做得出色同被人欣賞既機會亦會淢少 相反我相信行行出狀元 比心機做好一份你真係好有興趣既工 打工又好做生意又好 商機處處 千里馬又何懼搵唔倒伯樂 (伯樂都搵緊你架 )至於大公司同中小企呢方面 小弟亦有 d 心得 如果你係 d 穏穏陣陣求安定既人 搵間大 firm 啦 大 firm 通常個 hierarchy 同埋 bfnieit 都比細公司穏定 而且公司大 學好 d 球技都唔駛擔心會俾人炒 如果你係 d 雄心壯志要出人頭地既人 咁搵間 10-20 人既中小企都唔錯 因為人少 乜都要你做 自然學野既機會多 而且老闆亦會有多 d 機會知道你係乜水 咁對將來進升既機會又大 d 不過細公司有時真係有個種食君之祿擔君之休既感覺 見老闆幾個月都無生意個陣都咪話唔擔心 至於想創業既人事 一於學 CK 兄咁跳啦(我覺得係好事黎架) 由渣估兄所講既 2-4 人既菌企做到 26-50 人中小企就最好 因為一間得幾丁友既全場清道夫(啊四)管理模式 到 10 幾人份 team 同 30 幾 40 人份 department 既管理 怎至政治環境都有破大差別 打工時如果可以響老闆到有所領悟絶對比自己創業時先 大開眼介 著數
Add your comment below
Your Name:
Comment Title: